AI models

Every current AI model from the big labs, newest first: what each is for, its context size, its API price per million tokens and the product to use it in.

Crouton Alpha — stealth, November 2026. Lightweight model built for scalable coding, agentic, and multimodal workflows. (1M-token context)
GLM 5.3 FlashX Legal — Z.ai, November 2026. Z.ai's legal-domain variant of GLM-5.3-FlashX, a native multimodal model that accepts text, image, and video input and produces text. (1M-token context; API price $0.37 in and $1.25 out per 1M tokens)
AION 2 Cheat — AionLabs, October 2026. Aion 2: Full working installer for advanced AI prompt engineering tools, ready to run on your desktop. (open weights)
Nova V1.9.2 — Amazon, October 2026. Published: 2026-10-28. (open weights)
Particle-1 — Particle AI, October 2026. Token-efficient agentic coding model from Particle AI, post-trained on DeepSeek V4.1 Flash. (1M-token context; API price $0.05 in and $0.5 out per 1M tokens)
GPT-6.1 Terra — OpenAI, October 2026. Mid-tier model in OpenAI's GPT-6.1 series, priced below GPT-6.1 Sol. (1M-token context; API price $1 in and $5 out per 1M tokens)
GPT-6.1 Terra Pro — OpenAI, October 2026. Same underlying model as GPT-6.1 Terra, served with reasoning.mode set to pro for higher-quality responses on complex tasks. (1M-token context; API price $1 in and $5 out per 1M tokens)
Kimi K3.1 Pro — Moonshot AI, October 2026. Moonshot AI's flagship open-weight multimodal reasoning model and a post-trained successor to Kimi K3, with a 1M-token context window. (1M-token context; API price $2 in and $10 out per 1M tokens)
MiniMax M3.1 Flash — MiniMax, October 2026. Multimodal model from MiniMax. (1M-token context; API price $0.15 in and $0.6 out per 1M tokens; open weights)
DeepSeek V4.1 Flash 1017 — DeepSeek, October 2026. October 17, 2026 snapshot of DeepSeek V4.1 Flash, served by DeepSeek under the deepseek-flash API model name. (1M-token context; API price $0.3 in and $1.2 out per 1M tokens; open weights)
Glimpse — Elorian, October 2026. Multimodal visual reasoning model from Elorian, built for visual understanding and complex visual reasoning in the physical world. (262K-token context; API price $0.6 in and $3 out per 1M tokens)
Nemotron 3.5 Bolt — NVIDIA, October 2026. We are releasing NVIDIA Nemotron 3.5 Bolt 74B-A9B 24GB, a Mixture-of-Experts hybrid Mamba-Transformer model designed to run everyday agentic tasks locally on…
Beam 1.0 501B A23B — Reflection, October 2026. Reflection's open-weight reasoning model for coding and long-horizon agentic workflows. (1M-token context; API price $0.25 in and $1 out per 1M tokens; open weights)
Mercury 3 — Inception, October 2026. Diffusion-based language model from Inception, and the lab's most capable model to date. (1M-token context; API price $0.1 in and $0.35 out per 1M tokens)
Ternary Bonsai 2 9B — PrismML, October 2026. PrismML’s compact, open-weight language model for conversational AI, reasoning, and tool use. (262K-token context; API price $0.08 in and $0.13 out per 1M tokens; open weights)
Moondream 3 Preview — Moondream, October 2026. Vision-language model from Moondream with 9B total parameters and 2B active (mixture-of-experts). (32K-token context; API price $0.3 in and $2.5 out per 1M tokens; open weights)
Moondream 3.1 — Moondream, October 2026. Vision-language model from Moondream, the successor to Moondream 3 Preview, with 9B total parameters and 2B active (mixture-of-experts). (32K-token context; API price $0.3 in and $2.5 out per 1M tokens; open weights)
Muse Spark Open — Meta, October 2026. Coding-optimized open-weight model. (1M-token context; API price $1.25 in and $4.25 out per 1M tokens; open weights)
Nemotron 3.5 Super VL — NVIDIA, October 2026. NVIDIA Nemotron 3.5 Super VL is an open multimodal mixture-of-experts model from NVIDIA, with 12B active parameters out of 120B total. (1M-token context; API price $0.15 in and $0.5 out per 1M tokens; open weights)
Pareto 26.10 — unbiased, October 2026. Multimodal composite model built for research, coding, and agentic workflows, while delivering frontier-level performance across a broad range of… (1M-token context; API price $0.8 in and $3.2 out per 1M tokens)
Extract — ScaleDown, October 2026. Description-driven entity extraction model from ScaleDown. (1M-token context)
LFM2.5-24B-A2B — LiquidAI, October 2026. Mixture-of-experts reasoning model from Liquid AI with 24B total and 2.45B active parameters. (128K-token context; API price $0.05 in and $0.7 out per 1M tokens; open weights)
Pareto 26.9 — unbiased, October 2026. Multimodal composite model built for research, coding, and agentic workflows, while delivering frontier-level performance across a broad range of… (262K-token context; API price $2.5 in and $7.5 out per 1M tokens)
Summarize — ScaleDown, October 2026. Abstractive summarization model from ScaleDown. (1M-token context)
Muse Wonder — Meta, October 2026. Frontier reasoning model from Meta, designed for complex personal and professional agentic tasks. (1M-token context; API price $4 in and $20 out per 1M tokens)
Step 5 Preview — StepFun, October 2026. StepFun's flagship model for agentic work, built on a sparse Mixture-of-Experts architecture (27B active / 600B total parameters). (1M-token context; API price $1 in and $2.7 out per 1M tokens)
Claude Haiku 5.5 — Anthropic, October 2026. Anthropic's small, fast model for high-volume, cost-sensitive work such as summarization, subagents, and browser use. (1M-token context; API price $0.1 in and $0.5 out per 1M tokens)
Mistral Large 4 — Mistral, October 2026. Frontier multimodal (text and image input) model from Mistral AI built for reasoning, coding, and agentic workloads. (1M-token context; API price $0.7 in and $2.1 out per 1M tokens)
Nano Banana 2.1 — Google, October 2026. Google's image generation and editing model on the Flash tier, succeeding Nano Banana 2 and Nano Banana Pro. (65K-token context; API price $1.5 in and $7.5 out per 1M tokens)
Ling 3.1 Flash — inclusionAI, October 2026. Hybrid reasoning mixture-of-experts model from inclusionAI, with 25B active parameters out of 560B total. (1M-token context; API price $0.075 in and $0.225 out per 1M tokens)
Pareto 26.10 Preview — unbiased, October 2026. Multimodal composite model built for research, coding, and agentic workflows, while delivering frontier-level performance across a broad range of… (1M-token context; API price $0.8 in and $3.2 out per 1M tokens)
Gemini 4 Argon — Google, September 2026. Google’s frontier model for long, deep-reasoning work, with up to 1M tokens out; open only to vetted cyber-defence partners for now. (1M-token context; API price $2 in and $10 out per 1M tokens)
GPT-6.1 Sol — OpenAI, September 2026. An upgrade to GPT-6 Sol, the middle model of OpenAI's GPT-6 series. (1M-token context; API price $2 in and $10 out per 1M tokens)
GPT-6.1 Sol Pro — OpenAI, September 2026. GPT-6.1 Sol with deeper reasoning for complex tasks. (1M-token context; API price $2 in and $10 out per 1M tokens)
Claude Sonnet 5.5 — Anthropic, September 2026. Anthropic's everyday model for well-scoped work; a direct upgrade to Sonnet 5. (1M-token context; API price $2 in and $10 out per 1M tokens)
Jev Router — TypeSafe, September 2026. Jev Router picks the best model and reasoning effort for each request, balancing quality, speed, and cost. (1M-token context)
Perceptron Mk1.5 — Perceptron, September 2026. Perceptron's embodied reasoning model for physical agents. (36K-token context; API price $0.15 in and $1.5 out per 1M tokens)
Ember-1 — Fireworks, September 2026. Specialized reasoning model from Fireworks Research, built on Kimi K3. (1M-token context; API price $3 in and $15 out per 1M tokens)
Aion 3.5 — AionLabs, September 2026. Multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. (262K-token context; API price $3 in and $6 out per 1M tokens)
Aion 3.5 Mini — AionLabs, September 2026. Multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. (262K-token context; API price $0.7 in and $1.4 out per 1M tokens)
GLM 5.3 Prime — Z.ai, September 2026. GLM 5.3 served 1.5–2× faster. (1M-token context; API price $2.8 in and $8.8 out per 1M tokens)
Qwen3.8 Max Prime — Alibaba, September 2026. A faster, higher-priced version of Alibaba's Qwen3.8 Max. (1M-token context; API price $4 in and $12 out per 1M tokens)
Solar Mini 4 — Upstage, September 2026. Upstage's compact, cost-efficient language model, a 35B-parameter mixture-of-experts with 3B active parameters and a 524K context window. (524K-token context; API price $0.05 in and $0.2 out per 1M tokens)
Claude Opus 5.5 — Anthropic, September 2026. Anthropic's flagship for hard reasoning, coding and long agent tasks. (1M-token context; API price $4 in and $20 out per 1M tokens)
Command A+ — Cohere, September 2026. Cohere's flagship for business agent workflows. (256K-token context; API price $2.5 in and $10 out per 1M tokens)
GPT-6 Luna — OpenAI, September 2026. The fast, low-cost model in OpenAI's GPT-6 series. (1M-token context; API price $0.1 in and $0.5 out per 1M tokens)
GPT-6 Luna Pro — OpenAI, September 2026. Same underlying model as GPT-6 Luna, served with reasoning.mode set to pro for higher-quality responses on complex tasks. (1M-token context; API price $0.1 in and $0.5 out per 1M tokens)
GPT-6 Sol — OpenAI, September 2026. Cost-efficient high-end model in OpenAI's GPT-6 series, positioned below the flagship GPT-6 Astra and above the fast GPT-6 Luna tier. (1M-token context; API price $2 in and $10 out per 1M tokens)
GPT-6 Sol Pro — OpenAI, September 2026. Same underlying model as GPT-6 Sol, served with reasoning.mode set to pro for higher-quality responses on complex tasks. (1M-token context; API price $2 in and $10 out per 1M tokens)
Grok 4.7 — SpaceXAI, September 2026. SpaceXAI's flagship for coding, agent tasks and knowledge work. (500K-token context; API price $2 in and $6 out per 1M tokens)
MiMo-V2.6-Flash — Xiaomi, September 2026. Open-source foundation model developed by Xiaomi. (1M-token context; API price $0.14 in and $0.28 out per 1M tokens; open weights)
MiMo-V2.6-Pro — Xiaomi, September 2026. Flagship foundation model developed by Xiaomi. (1M-token context; API price $0.435 in and $0.87 out per 1M tokens; open weights)
MiMo-V2.6-Pro-UltraSpeed — Xiaomi, September 2026. Fast speed edition of Xiaomi's flagship foundation model, MiMo-V2.6-Pro. (1M-token context; API price $4.35 in and $8.7 out per 1M tokens)
Qwen3.8 Omni Flash — Alibaba, September 2026. Alibaba’s agent-first model that understands audio and video. (1M-token context; API price $0.15 in and $0.47 out per 1M tokens)
Switchyard — NVIDIA, September 2026. Open-source model router that switches between multiple models to optimize the cost of requests. (1M-token context)
GLM 5.3 FlashX — Z.ai, September 2026. High-speed variant of Z.ai's GLM-5.3-Flash, a native multimodal model delivering inference speeds of up to 200 tokens/s. (1M-token context; API price $0.37 in and $1.25 out per 1M tokens)
Ternary Bonsai 2 27B — PrismML, September 2026. Bonsai 2 27B is a 27B-parameter reasoning model from PrismML derived from Qwen3.8-27B. (262K-token context; API price $0.075 in and $0.5 out per 1M tokens; open weights)
Pareto — unbiased, September 2026. Multimodal composite model built for research, coding, and agentic workflows, while delivering frontier-level performance across a broad range of… (262K-token context; API price $2.5 in and $7.5 out per 1M tokens)
Schematron V2 Small — Inference.net, September 2026. 3B-parameter HTML-to-JSON extraction model from Inference.net. (128K-token context; API price $0.05 in and $0.23 out per 1M tokens; open weights)
Schematron V2 Turbo — Inference.net, September 2026. 3B-parameter HTML-to-JSON extraction model from Inference.net. (128K-token context; API price $0.03 in and $0.15 out per 1M tokens; open weights)
Fugu Max — Sakana, September 2026. Cost-performance model in Sakana AI's Fugu family. (1M-token context; API price $2 in and $6 out per 1M tokens)
Fugu Ultra v2 — Sakana, September 2026. Higher-performance model in Sakana AI's Fugu family. (1M-token context; API price $5 in and $30 out per 1M tokens)
DeepSeek V4.1 Flash — DeepSeek, September 2026. DeepSeek's fast open model, on a new architecture. (1M-token context; API price $0.045 in and $0.5 out per 1M tokens; open weights)
Ling 3.0 Flash VL — inclusionAI, September 2026. Ling 3.0 Flash VL builds on Ling 3.0 Flash (124B total / 5.5B active MoE from InclusionAI), further strengthening its language capabilities while adding… (131K-token context; API price $0.06 in and $0.18 out per 1M tokens; open weights)
Mercury 2.5 — Inception, September 2026. Fastest reasoning LLM, and the latest diffusion LLM (dLLM) from Inception. (260K-token context; API price $0.04 in and $0.15 out per 1M tokens)
Nex-N2.5-Mini — Nex AGI, September 2026. Nex-N2.5 is an agentic model built to turn goals into working, verified outcomes. (262K-token context; API price $0.025 in and $0.1 out per 1M tokens; open weights)
Nex-N2.5-Pro — Nex AGI, September 2026. Nex-N2.5 is an agentic model built to turn goals into working, verified outcomes. (262K-token context; API price $0.075 in and $0.25 out per 1M tokens; open weights)
GPT-6 Astra — OpenAI, September 2026. OpenAI's flagship model for demanding end-to-end work. (1M-token context; API price $10 in and $50 out per 1M tokens)
GPT-6 Astra Pro — OpenAI, September 2026. Same underlying model as GPT-6 Astra, served with reasoning.mode set to pro for higher-quality responses on complex tasks. (1M-token context; API price $10 in and $50 out per 1M tokens)
Ling 3.0 Flash Sante — inclusionAI, September 2026. Health and medicine-focused mixture-of-experts model from InclusionAI, built on Ling 3.0 Flash with 5.1B active parameters out of 124B total. (262K-token context; API price $0.042 in and $0.123 out per 1M tokens)
Qwen3.8 Max (0902) — Alibaba, September 2026. Updated snapshot of Qwen3.8 Max from Alibaba's Qwen team. (1M-token context; API price $2 in and $6 out per 1M tokens)
Gemini 3.8 Flash — Google, September 2026. Google's smartest Flash model, better at coding, agents and reasoning. (1M-token context; API price $0.825 in and $4.13 out per 1M tokens)
Muse Spark 1.3 — Meta, September 2026. Multimodal reasoning model from Meta for long-running agentic, multi-agent, and coding workflows. (1M-token context; API price $1.25 in and $4.25 out per 1M tokens)
Muse Spark 1.3 Contributor — Meta, September 2026. Cost-efficient contributor tier of Meta’s multimodal reasoning model for experimentation, learning, and early-stage agentic, multi-agent, and coding workflows. (1M-token context; API price $0.1 in and $0.2 out per 1M tokens)
Claude Fable 5.1 — Anthropic, September 2026. Improves on Fable 5, most of all in agentic coding and long-running work. (1M-token context; API price $10 in and $50 out per 1M tokens)
Granite 4.2 8B — IBM, August 2026. Dense reasoning model from IBM. (131K-token context; API price $0.06 in and $0.25 out per 1M tokens; open weights)
Hy4 preview — Tencent, August 2026. Tencent: Hy4 preview is a mixture-of-experts model from Tencent, with 49B active parameters out of 770B total. (1M-token context; API price $0.834 in and $2.5 out per 1M tokens; open weights)
Ling 3.0 Flash Fin — inclusionAI, August 2026. Finance-focused mixture-of-experts model from InclusionAI, built on Ling 3.0 Flash with 5.1B active parameters out of 124B total. (262K-token context; API price $0.042 in and $0.123 out per 1M tokens)
GLM 5.3 Flash — Z.ai, August 2026. Native multimodal model from Z.ai. (1M-token context; API price $0.031 in and $0.085 out per 1M tokens; open weights)
Qwen3.8 Flash — Alibaba, August 2026. Multimodal reasoning model from Alibaba. (1M-token context; API price $0.15 in and $0.47 out per 1M tokens; open weights)
DeepSeek V4 Flash Vision Exp — DeepSeek, August 2026. Experimental vision-enabled version of DeepSeek V4 Flash 0731 from DeepSeek, adding image understanding while matching the base model on text capabilities… (1M-token context; API price $0.216 in and $0.647 out per 1M tokens; open weights)
Muse Spark 1.2 Contributor — Meta, August 2026. Reasoning model from Meta designed for developers who want to start building at an even lower cost. (1M-token context; API price $0.1 in and $0.2 out per 1M tokens)
Hy-MT2-1.8B — Tencent, August 2026. Compact 1.8B-parameter translation model from Tencent. (8K-token context; API price $0.044 in and $0.177 out per 1M tokens; open weights)
Hy-MT2-30B-A3B — Tencent, August 2026. Tencent's flagship translation model in the Hy-MT2 family. (8K-token context; API price $0.074 in and $0.295 out per 1M tokens; open weights)
Hy-MT2-7B — Tencent, August 2026. 7B-parameter translation model from Tencent. (8K-token context; API price $0.074 in and $0.295 out per 1M tokens; open weights)
GLM 5.3 — Z.ai, August 2026. Z.ai's open reasoning model for software engineering and long agent tasks. (1M-token context; API price $0.08 in and $3.6 out per 1M tokens; open weights)
Qwen3.8 27B — Alibaba, August 2026. An open model from Alibaba's Qwen team that reads text and images. (1M-token context; API price $0.15 in and $1.88 out per 1M tokens; open weights)
Gemini 3.7 Flash — Google, August 2026. Fast multimodal model for agents, coding and multi-step reasoning. (1M-token context; API price $0.75 in and $3.75 out per 1M tokens)
DeepSeek V4 Pro 0813 — DeepSeek, August 2026. Large-scale mixture-of-experts model from DeepSeek. (1M-token context; API price $0.115 in and $4 out per 1M tokens; open weights)
Grok 4.6 — SpaceXAI, August 2026. The previous Grok: strong at coding, knowledge work and STEM. (500K-token context; API price $2 in and $6 out per 1M tokens)
Qwen3.8 2.4T A95B — Alibaba, August 2026. Open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of Qwen3.8 Max, with 95 billion active parameters out of 2.4 trillion total. (1M-token context; API price $2 in and $6 out per 1M tokens; open weights)
Seed 2.1 Turbo — ByteDance Seed, August 2026. Multimodal model from ByteDance Seed for coding and long-horizon agent workflows. (262K-token context; API price $0.5 in and $2.5 out per 1M tokens)
LFM2.5-2.6B — LiquidAI, August 2026. Compact reasoning model from Liquid AI. (128K-token context; API price $0.35 in and $0.5 out per 1M tokens; open weights)
Nemotron 3.5 Lightning — NVIDIA, August 2026. NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model from NVIDIA, with 3B active parameters out of 30B total. (262K-token context; API price $0.07 in and $0.2 out per 1M tokens; open weights)
Sakana Namazu — Sakana, August 2026. Japanese-specialized reasoning model from Sakana AI, based on Kimi K2.6 with additional training for Japanese language and business contexts. (262K-token context; API price $0.95 in and $4 out per 1M tokens)
Solar Pro 4 — Upstage, August 2026. Upstage's cost-efficient large language model, featuring a 524K context window. (524K-token context; API price $0.15 in and $0.6 out per 1M tokens)
Muse Glimmer 30B — Meta, August 2026. Dense, open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark and optimized for autonomous agents on consumer hardware. (131K-token context; API price $0.3 in and $1.2 out per 1M tokens; open weights)
Muse Spark 1.2 — Meta, August 2026. Reasoning model from Meta, designed for complex agentic tasks. (1M-token context; API price $1.25 in and $4.25 out per 1M tokens)
DeepSeek V4 Flash 0731 — DeepSeek, July 2026. Sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. (1M-token context; API price $0.059 in and $0.3 out per 1M tokens; open weights)
Inkling Small — Thinking Machines, July 2026. Open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 12B active parameters out of 276B total. (524K-token context; API price $0.45 in and $1.2 out per 1M tokens; open weights)
MiniMax H3 — MiniMax, July 2026. MiniMax’s open omni-modal model that generates video with stereo audio, up to 2K and 15 seconds, from frames or references. (open weights)
Qwen3.7 Flash — Alibaba, July 2026. Vision-language reasoning model from Alibaba. (1M-token context; API price $0.03 in and $0.13 out per 1M tokens)
Claude Opus 5 — Anthropic, July 2026. The previous Opus: hard reasoning, coding and long agent tasks. (1M-token context; API price $5 in and $25 out per 1M tokens)
Ling 3.0 Flash — inclusionAI, July 2026. 124B-parameter Mixture-of-Experts (MoE) model, with approximately 5.1B parameters activated per token. (262K-token context; API price $0.021 in and $0.063 out per 1M tokens; open weights)
Gemini 3.5 Flash Lite — Google, July 2026. Google's high-efficiency model, with better agent skills. (1M-token context; API price $0.3 in and $2.5 out per 1M tokens)
Gemini 3.6 Flash — Google, July 2026. High-efficiency model from Google for coding, agentic workflows, and web and app development. (1M-token context; API price $0.75 in and $3.75 out per 1M tokens)
LongCat 2.0 — Meituan, July 2026. Sparse mixture-of-experts language model from Meituan, with 48B active parameters out of 1.6T total. (1M-token context; API price $0.3 in and $1.2 out per 1M tokens; open weights)
Inkling — Thinking Machines, July 2026. Open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 41B active parameters out of 975B total. (524K-token context; API price $1 in and $4.05 out per 1M tokens; open weights)
Kimi K3 — Moonshot AI, July 2026. Moonshot AI's very large open multimodal reasoning model. (1M-token context; API price $0.2 in and $16 out per 1M tokens; open weights)
Muse Spark 1.1 — Meta, July 2026. Multimodal reasoning model from Meta, built for agentic tasks. (1M-token context; API price $1.25 in and $4.25 out per 1M tokens)
GPT-5.6 Luna — OpenAI, July 2026. Fast, cost-efficient model in OpenAI's GPT-5.6 series. (1M-token context; API price $0.2 in and $1.2 out per 1M tokens)
GPT-5.6 Luna Pro — OpenAI, July 2026. Same underlying model as GPT-5.6 Luna, served with reasoning.mode set to pro for higher-quality responses on complex tasks. (1M-token context; API price $0.2 in and $1.2 out per 1M tokens)
GPT-5.6 Sol — OpenAI, July 2026. Flagship model in OpenAI's GPT-5.6 series. (1M-token context; API price $4 in and $20 out per 1M tokens)
GPT-5.6 Sol Pro — OpenAI, July 2026. Same underlying model as GPT-5.6 Sol, served with reasoning.mode set to pro for higher-quality responses on complex tasks. (1M-token context; API price $4 in and $20 out per 1M tokens)
GPT-5.6 Terra — OpenAI, July 2026. Balanced model in OpenAI's GPT-5.6 series, positioned between the flagship Sol tier and the cost-efficient Luna tier. (1M-token context; API price $2 in and $12 out per 1M tokens)
GPT-5.6 Terra Pro — OpenAI, July 2026. Same underlying model as GPT-5.6 Terra, served with reasoning.mode set to pro for higher-quality responses on complex tasks. (1M-token context; API price $2 in and $12 out per 1M tokens)
Grok 4.5 — SpaceXAI, July 2026. Model from SpaceXAI with frontier performance on coding, knowledge work, and STEM. (500K-token context; API price $2 in and $6 out per 1M tokens)
Aion-3.0 — AionLabs, July 2026. Multi-model roleplaying and storytelling system from AionLabs, built on the GLM family of models. (131K-token context; API price $3 in and $6 out per 1M tokens)
Aion-3.0-Mini — AionLabs, July 2026. Aion-3.0 Mini is a multi-model roleplaying and storytelling system from AionLabs, built on the DeepSeek family of models. (131K-token context; API price $0.7 in and $1.4 out per 1M tokens)
Hy3 — Tencent, July 2026. 295B-parameter Mixture-of-Experts model from Tencent (21B active, 192 experts with top-8 routing) built for reasoning, agentic workflows, and real-world… (262K-token context; API price $0.132 in and $0.528 out per 1M tokens; open weights)
Claude Sonnet 5 — Anthropic, June 2026. The previous Sonnet: strong at coding, agents and professional work. (1M-token context; API price $2 in and $10 out per 1M tokens)
Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) — Google, June 2026. Google's fastest, most cost-efficient Gemini image model, built for high-velocity developer pipelines and rapid-fire visual exploration. (65K-token context; API price $0.25 in and $1.5 out per 1M tokens)
Fugu Ultra — Sakana, June 2026. Higher-performance model in Sakana AI's Fugu family. (1M-token context; API price $5 in and $30 out per 1M tokens)
Nano Banana 2 — Google, June 2026. Google's image generation and editing model (Gemini 3.1 Flash Image). (131K-token context; API price $0.5 in and $3 out per 1M tokens)
Nano Banana Pro (Gemini 3 Pro Image) — Google, June 2026. Google’s most advanced image-generation and editing model, built on Gemini 3 Pro. (131K-token context; API price $2 in and $12 out per 1M tokens)
GLM 5.2 — Z.ai, June 2026. Large-scale reasoning model from Z.ai. (1M-token context; API price $0.006 in and $6.8 out per 1M tokens; open weights)
Kimi K2.7 Code — Moonshot AI, June 2026. MoonshotAI: Kimi K2.7 Code is a coding-focused model in Moonshot AI's Kimi K2 family, built to complete end-to-end programming tasks reliably over long… (262K-token context; API price $0.6 in and $3.5 out per 1M tokens; open weights)
Claude Fable 5 — Anthropic, June 2026. Mythos-class model from Anthropic, built for autonomous knowledge work and coding. (1M-token context; API price $10 in and $50 out per 1M tokens)
Nemotron 3 Ultra — NVIDIA, June 2026. NVIDIA's open model for frontier reasoning and orchestration. (262K-token context; API price $0.6 in and $2.4 out per 1M tokens; open weights)
Nemotron 3.5 Content Safety — NVIDIA, June 2026. NVIDIA Nemotron 3.5 Content Safety is a compact 4B-parameter multimodal guardrail model from NVIDIA, fine-tuned from Google Gemma-3-4B. (131K-token context; API price $0.2 in and $0.2 out per 1M tokens; open weights)
Qwen3.7 Plus — Alibaba, June 2026. Cost-effective model in Alibaba's Qwen3.7 series. (1M-token context; API price $0.32 in and $1.28 out per 1M tokens)
MiniMax M3 — MiniMax, May 2026. MiniMax's open multimodal foundation model. (1M-token context; API price $0.3 in and $1.2 out per 1M tokens; open weights)
Step 3.7 Flash — StepFun, May 2026. StepFun's latest high-efficiency multimodal Mixture-of-Experts model. (262K-token context; API price $0.2 in and $1.15 out per 1M tokens; open weights)
Claude Opus 4.8 — Anthropic, May 2026. Anthropic's most capable generally available model in the Opus family. (1M-token context; API price $5 in and $25 out per 1M tokens)
Qwen3.7 Max — Alibaba, May 2026. Flagship model in Alibaba's Qwen3.7 series. (1M-token context; API price $1.48 in and $4.42 out per 1M tokens)
Grok Build 0.1 — SpaceXAI, May 2026. SpaceXAI’s fast coding model trained specifically for agentic software engineering workflows. (256K-token context; API price $1 in and $2 out per 1M tokens)
Gemini 3.5 Flash — Google, May 2026. Google's high-efficiency multimodal model, bringing near-Pro level coding and reasoning at Flash-tier cost and speed. (1M-token context; API price $1.5 in and $9 out per 1M tokens)
Perceptron Mk1 — Perceptron, May 2026. Perceptron's highest-quality vision-language model for video and embodied reasoning. (32K-token context; API price $0.15 in and $1.5 out per 1M tokens)
Gemini 3.1 Flash Lite — Google, May 2026. Google’s GA high-efficiency multimodal model optimized for low-latency, high-volume workloads. (1M-token context; API price $0.25 in and $1.5 out per 1M tokens)
GPT Chat Latest — OpenAI, May 2026. GPT Chat Latest points to OpenAI's stable API alias chat-latest that always resolves to the latest Instant chat model used in ChatGPT. (400K-token context; API price $5 in and $30 out per 1M tokens)
Grok 4.3 — SpaceXAI, April 2026. Reasoning model from SpaceXAI. (1M-token context; API price $1.25 in and $2.5 out per 1M tokens)
Mistral Medium 3.5 — Mistral, April 2026. Mistral's dense 128B model for following instructions. (262K-token context; API price $1.5 in and $7.5 out per 1M tokens)
Qwen3.5 Plus 2026-04-20 — Alibaba, April 2026. Large-scale multimodal language model from Alibaba. (1M-token context; API price $0.3 in and $1.8 out per 1M tokens)
Qwen3.6 27B — Alibaba, April 2026. Dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026. (262K-token context; API price $0.45 in and $2.7 out per 1M tokens; open weights)
Qwen3.6 35B A3B — Alibaba, April 2026. Open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per token. (262K-token context; API price $0.1 in and $0.9 out per 1M tokens; open weights)
Qwen3.6 Flash — Alibaba, April 2026. Fast, efficient language model from Alibaba's Qwen 3.6 series. (1M-token context; API price $0.188 in and $1.13 out per 1M tokens)
DeepSeek V4 Flash 0423 — DeepSeek, April 2026. Efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. (1M-token context; API price $0.091 in and $0.182 out per 1M tokens; open weights)
DeepSeek V4 Pro — DeepSeek, April 2026. DeepSeek's large open model with a 1M-token context. (1M-token context; API price $1.6 in and $3.2 out per 1M tokens; open weights)
GPT-5.5 — OpenAI, April 2026. OpenAI’s frontier model designed for complex professional workloads, building on GPT-5.4 with stronger reasoning, higher reliability, and improved token… (1M-token context; API price $5 in and $30 out per 1M tokens)
GPT-5.5 Pro — OpenAI, April 2026. OpenAI’s high-capability model optimized for deep reasoning and accuracy on complex, high-stakes workloads. (1M-token context; API price $30 in and $180 out per 1M tokens)
Hy3 preview — Tencent, April 2026. High-efficiency Mixture-of-Experts model from Tencent designed for agentic workflows and production use. (262K-token context; API price $0.18 in and $0.6 out per 1M tokens; open weights)
MiMo-V2.5 — Xiaomi, April 2026. Native omnimodal model by Xiaomi. (1M-token context; API price $0.4 in and $2 out per 1M tokens; open weights)
MiMo-V2.5-Pro — Xiaomi, April 2026. Xiaomi’s flagship model, delivering strong performance in general agentic capabilities, complex software engineering, and long-horizon tasks, with top… (1M-token context; API price $0.48 in and $1.8 out per 1M tokens; open weights)
GPT-5.4 Image 2 — OpenAI, April 2026. GPT-5.4 Image 2 combines OpenAI's GPT-5.4 model with state-of-the-art image generation capabilities from GPT Image 2. (272K-token context; API price $8 in and $15 out per 1M tokens)
Kimi K2.6 — Moonshot AI, April 2026. Moonshot AI's next-generation multimodal model, designed for long-horizon coding, coding-driven UI/UX generation, and multi-agent orchestration. (262K-token context; API price $0.65 in and $3.41 out per 1M tokens; open weights)
Claude Opus 4.7 — Anthropic, April 2026. Opus 4.7 is the next generation of Anthropic's Opus family, built for long-running, asynchronous agents. (1M-token context; API price $5 in and $25 out per 1M tokens)
GLM 5.1 — Z.ai, April 2026. GLM-5.1 delivers a major leap in coding capability, with particularly significant gains in handling long-horizon tasks. (204K-token context; API price $1.4 in and $4.4 out per 1M tokens; open weights)
Gemma 4 26B A4B — Google, April 2026. Instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. (262K-token context; API price $0.076 in and $0.228 out per 1M tokens; open weights)
Gemma 4 31B — Google, April 2026. Google DeepMind's open model that reads text and images. (262K-token context; API price $0.2 in and $0.4 out per 1M tokens; open weights)
Qwen3.6 Plus — Alibaba, April 2026. Qwen 3.6 Plus builds on a hybrid architecture that combines efficient linear attention with sparse mixture-of-experts routing, enabling strong scalability and… (1M-token context; API price $0.325 in and $1.95 out per 1M tokens)
GLM 5V Turbo — Z.ai, April 2026. Z.ai’s first native multimodal agent foundation model, built for vision-based coding and agent-driven tasks. (202K-token context; API price $1.2 in and $4 out per 1M tokens)
Trinity Large Thinking — Arcee AI, April 2026. Powerful open source reasoning model from the team at Arcee AI. (262K-token context; API price $0.25 in and $0.8 out per 1M tokens; open weights)
Grok 4.20 — SpaceXAI, March 2026. Reasoning model from SpaceXAI with industry-leading speed and agentic tool calling capabilities. (2M-token context; API price $1.25 in and $2.5 out per 1M tokens)
Grok 4.20 Multi-Agent — SpaceXAI, March 2026. Variant of SpaceXAI’s Grok 4.20 designed for collaborative, agent-based workflows. (2M-token context; API price $1.25 in and $2.5 out per 1M tokens)
Lyria 3 Clip Preview — Google, March 2026. 30 second duration clips are priced at $0.04 per clip. (1M-token context)
Lyria 3 Pro Preview — Google, March 2026. Full-length songs are priced at $0.08 per song. (1M-token context)
Reka Edge — rekaai, March 2026. Extremely efficient 7B multimodal vision-language model that accepts image/video+text inputs and generates text outputs. (16K-token context; API price $0.1 in and $0.1 out per 1M tokens; open weights)
MiniMax M2.7 — MiniMax, March 2026. Next-generation large language model designed for autonomous, real-world productivity and continuous improvement. (204K-token context; API price $0.3 in and $1.2 out per 1M tokens; open weights)
GPT-5.4 Mini — OpenAI, March 2026. GPT-5.4 mini brings the core capabilities of GPT-5.4 to a faster, more efficient model optimized for high-throughput workloads. (400K-token context; API price $0.75 in and $4.5 out per 1M tokens)
GPT-5.4 Nano — OpenAI, March 2026. Most lightweight and cost-efficient variant of the GPT-5.4 family, optimized for speed-critical and high-volume tasks. (400K-token context; API price $0.2 in and $1.25 out per 1M tokens)
Mistral Small 4 — Mistral, March 2026. Mistral's open small model that combines several flagship skills. (262K-token context; API price $0.15 in and $0.6 out per 1M tokens; open weights)
GLM 5 Turbo — Z.ai, March 2026. New model from Z.ai designed for fast inference and strong performance in agent-driven environments such as OpenClaw scenarios. (202K-token context; API price $1.2 in and $4 out per 1M tokens)
Nemotron 3 Super — NVIDIA, March 2026. NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex… (1M-token context; API price $0.3 in and $2.4 out per 1M tokens; open weights)
Qwen3.5-9B — Alibaba, March 2026. Multimodal foundation model from the Qwen3.5 family, designed to deliver strong reasoning, coding, and visual understanding in an efficient 9B-parameter… (262K-token context; API price $0.17 in and $0.25 out per 1M tokens; open weights)
Seed-2.0-Lite — ByteDance Seed, March 2026. Versatile, cost‑efficient enterprise workhorse that delivers strong multimodal and agent capabilities while offering noticeably lower latency, making it a… (262K-token context; API price $0.25 in and $2 out per 1M tokens)
GPT-5.4 — OpenAI, March 2026. OpenAI’s latest frontier model, unifying the Codex and GPT lines into a single system. (1M-token context; API price $2.5 in and $15 out per 1M tokens)
GPT-5.4 Pro — OpenAI, March 2026. OpenAI's most advanced model, building on GPT-5.4's unified architecture with enhanced reasoning capabilities for complex, high-stakes tasks. (1M-token context; API price $30 in and $180 out per 1M tokens)
Mercury 2 — Inception, March 2026. Extremely fast reasoning LLM, and the first reasoning diffusion LLM (dLLM). (128K-token context; API price $0.25 in and $0.75 out per 1M tokens)
Gemini 3.1 Flash Lite Preview — Google, March 2026. Google's high-efficiency model optimized for high-volume use cases. (1M-token context; API price $0.25 in and $1.5 out per 1M tokens)
Nano Banana 2 (Gemini 3.1 Flash Image Preview) — Google, February 2026. Gemini 3.1 Flash Image Preview, a.k.a. (65K-token context; API price $0.5 in and $3 out per 1M tokens)
Seed-2.0-Mini — ByteDance Seed, February 2026. Seed-2.0-mini targets latency-sensitive, high-concurrency, and cost-sensitive scenarios, emphasizing fast response and flexible inference deployment. (262K-token context; API price $0.1 in and $0.4 out per 1M tokens)
Gemini 3.1 Pro Preview Custom Tools — Google, February 2026. Variant of Gemini 3.1 Pro that improves tool selection behavior by preventing overuse of a general bash tool when more efficient third-party... (1M-token context; API price $2 in and $12 out per 1M tokens)
Qwen3.5-122B-A10B — Alibaba, February 2026. The Qwen3.5 122B-A10B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse… (262K-token context; API price $0.26 in and $2.08 out per 1M tokens; open weights)
Qwen3.5-27B — Alibaba, February 2026. The Qwen3.5 27B native vision-language Dense model incorporates a linear attention mechanism, delivering fast response times while balancing inference speed… (262K-token context; API price $0.26 in and $2.6 out per 1M tokens; open weights)
Qwen3.5-35B-A3B — Alibaba, February 2026. The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse… (262K-token context; API price $0.14 in and $1 out per 1M tokens; open weights)
Qwen3.5-Flash — Alibaba, February 2026. The Qwen3.5 native vision-language Flash models are built on a hybrid architecture that integrates a linear attention mechanism with a sparse… (1M-token context; API price $0.065 in and $0.26 out per 1M tokens)
GPT-5.3-Codex — OpenAI, February 2026. OpenAI’s most advanced agentic coding model, combining the frontier software engineering performance of GPT-5.2-Codex with the broader reasoning and… (400K-token context; API price $1.75 in and $14 out per 1M tokens)
Aion-2.0 — AionLabs, February 2026. Variant of DeepSeek V3.2 optimized for immersive roleplaying and storytelling. (131K-token context; API price $0.8 in and $1.6 out per 1M tokens)
Gemini 3.1 Pro Preview — Google, February 2026. Google’s frontier reasoning model, delivering enhanced software engineering performance, improved agentic reliability, and more efficient token usage across… (1M-token context; API price $2 in and $12 out per 1M tokens)
Claude Sonnet 4.6 — Anthropic, February 2026. Sonnet 4.6 is Anthropic's most capable Sonnet-class model yet, with frontier performance across coding, agents, and professional work. (1M-token context; API price $3 in and $15 out per 1M tokens)
Qwen3.5 397B A17B — Alibaba, February 2026. The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse… (262K-token context; API price $0.5 in and $3.6 out per 1M tokens; open weights)
Qwen3.5 Plus 2026-02-15 — Alibaba, February 2026. The Qwen3.5 native vision-language series Plus models are built on a hybrid architecture that integrates linear attention mechanisms with sparse… (1M-token context; API price $0.26 in and $1.56 out per 1M tokens)
MiniMax M2.5 — MiniMax, February 2026. SOTA large language model designed for real-world productivity. (204K-token context; API price $0.27 in and $0.95 out per 1M tokens; open weights)
GLM 5 — Z.ai, February 2026. Z.ai’s flagship open-source foundation model engineered for complex systems design and long-horizon agent workflows. (204K-token context; API price $1 in and $3.2 out per 1M tokens; open weights)
Claude Opus 4.6 — Anthropic, February 2026. Opus 4.6 is Anthropic’s strongest model for coding and long-running professional tasks. (1M-token context; API price $5 in and $25 out per 1M tokens)
Qwen3 Coder Next — Alibaba, February 2026. Open-weight causal language model optimized for coding agents and local development workflows. (262K-token context; API price $0.12 in and $0.8 out per 1M tokens; open weights)
Step 3.5 Flash — StepFun, January 2026. StepFun's most capable open-source foundation model. (262K-token context; API price $0.1 in and $0.3 out per 1M tokens; open weights)
Kimi K2.5 — Moonshot AI, January 2026. Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm. (262K-token context; API price $0.5 in and $2.5 out per 1M tokens; open weights)
Solar Pro 3 — Upstage, January 2026. Upstage's powerful Mixture-of-Experts (MoE) language model. (131K-token context; API price $0.15 in and $0.6 out per 1M tokens)
MiniMax M2-her — MiniMax, January 2026. Dialogue-first large language model built for immersive roleplay, character-driven chat, and expressive multi-turn conversations. (65K-token context; API price $0.3 in and $1.2 out per 1M tokens)
Palmyra X5 — Writer, January 2026. Writer's most advanced model, purpose-built for building and scaling AI agents across the enterprise. (1M-token context; API price $0.6 in and $6 out per 1M tokens)
GLM 4.7 Flash — Z.ai, January 2026. As a 30B-class SOTA model, GLM-4.7-Flash offers a new option that balances performance and efficiency. (200K-token context; API price $0.06 in and $0.4 out per 1M tokens; open weights)
GPT Audio — OpenAI, January 2026. The gpt-audio model is OpenAI's first generally available audio model. (128K-token context; API price $2.5 in and $10 out per 1M tokens)
GPT Audio Mini — OpenAI, January 2026. A cost-efficient version of GPT Audio. (128K-token context; API price $0.6 in and $2.4 out per 1M tokens)
GPT-5.2-Codex — OpenAI, January 2026. Upgraded version of GPT-5.1-Codex optimized for software engineering and coding workflows. (400K-token context; API price $1.75 in and $14 out per 1M tokens)
MiniMax M2.1 — MiniMax, December 2025. Lightweight, state-of-the-art large language model optimized for coding, agentic workflows, and modern application development. (204K-token context; API price $0.3 in and $1.2 out per 1M tokens; open weights)
GLM 4.7 — Z.ai, December 2025. Z.ai’s latest flagship model, featuring upgrades in two key areas: enhanced programming capabilities and more stable multi-step reasoning/execution. (204K-token context; API price $0.54 in and $1.98 out per 1M tokens; open weights)
Gemini 3 Flash Preview — Google, December 2025. High speed, high value thinking model designed for agentic workflows, multi turn chat, and coding assistance. (1M-token context; API price $0.5 in and $3 out per 1M tokens)
Nemotron 3 Nano 30B A3B — NVIDIA, December 2025. NVIDIA Nemotron 3 Nano 30B A3B is a small language MoE model with highest compute efficiency and accuracy for developers to build specialized agentic AI… (262K-token context; API price $0.06 in and $0.24 out per 1M tokens; open weights)
GPT-5.2 — OpenAI, December 2025. Latest frontier-grade model in the GPT-5 series, offering stronger agentic and long context perfomance compared to GPT-5.1. (400K-token context; API price $1.75 in and $14 out per 1M tokens)
GPT-5.2 Chat — OpenAI, December 2025. Fast, lightweight member of the 5.2 family, optimized for low-latency chat while retaining strong general intelligence. (128K-token context; API price $1.75 in and $14 out per 1M tokens)
GPT-5.2 Pro — OpenAI, December 2025. OpenAI’s most advanced model, offering major improvements in agentic coding and long context performance over GPT-5 Pro. (400K-token context; API price $21 in and $168 out per 1M tokens)
Devstral 2 — Mistral, December 2025. Mistral's open model for agentic coding. (262K-token context; API price $0.4 in and $2 out per 1M tokens; open weights)
GLM 4.6V — Z.ai, December 2025. Large multimodal model designed for high-fidelity visual understanding and long-context reasoning across images, documents, and mixed media. (131K-token context; API price $0.3 in and $0.9 out per 1M tokens; open weights)
Relace Search — Relace, December 2025. The relace-search model uses 4-12 viewfile and grep tools in parallel to explore a codebase and return relevant files to the user request. (256K-token context; API price $1 in and $3 out per 1M tokens)
GPT-5.1-Codex-Max — OpenAI, December 2025. OpenAI’s latest agentic coding model, designed for long-running, high-context software development tasks. (400K-token context; API price $1.25 in and $10 out per 1M tokens)
Ministral 3 14B 2512 — Mistral, December 2025. The largest model in the Ministral 3 family, Ministral 3 14B offers frontier capabilities and performance comparable to its larger Mistral Small 3.2 24B… (262K-token context; API price $0.2 in and $0.2 out per 1M tokens; open weights)
Ministral 3 3B 2512 — Mistral, December 2025. The smallest model in the Ministral 3 family, Ministral 3 3B is a powerful, efficient tiny language model with vision capabilities. (131K-token context; API price $0.1 in and $0.1 out per 1M tokens; open weights)
Ministral 3 8B 2512 — Mistral, December 2025. A balanced model in the Ministral 3 family, Ministral 3 8B is a powerful, efficient tiny language model with vision capabilities. (262K-token context; API price $0.15 in and $0.15 out per 1M tokens; open weights)
Nova 2 Lite — Amazon, December 2025. Fast, cost-effective reasoning model for everyday workloads that can process text, images, and videos to generate text. (1M-token context; API price $0.3 in and $2.5 out per 1M tokens)
DeepSeek V3.2 — DeepSeek, December 2025. Large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. (163K-token context; API price $0.26 in and $0.38 out per 1M tokens; open weights)
Mistral Large 3 2512 — Mistral, December 2025. Mistral’s most capable model to date, featuring a sparse mixture-of-experts architecture with 41B active parameters (675B total), and released under the… (262K-token context; API price $0.5 in and $1.5 out per 1M tokens)
Claude Opus 4.5 — Anthropic, November 2025. Anthropic’s frontier reasoning model optimized for complex software engineering, agentic workflows, and long-horizon computer use. (200K-token context; API price $5 in and $25 out per 1M tokens)
Nano Banana Pro (Gemini 3 Pro Image Preview) — Google, November 2025. Google’s most advanced image-generation and editing model, built on Gemini 3 Pro. (65K-token context; API price $2 in and $12 out per 1M tokens)
GPT-5.1 — OpenAI, November 2025. Latest frontier-grade model in the GPT-5 series, offering stronger general-purpose reasoning, improved instruction adherence, and a more natural… (400K-token context; API price $1.25 in and $10 out per 1M tokens)
GPT-5.1-Codex — OpenAI, November 2025. Specialized version of GPT-5.1 optimized for software engineering and coding workflows. (400K-token context; API price $1.25 in and $10 out per 1M tokens)
GPT-5.1-Codex-Mini — OpenAI, November 2025. Smaller and faster version of GPT-5.1-Codex. (400K-token context; API price $0.25 in and $2 out per 1M tokens)
Kimi K2 Thinking — Moonshot AI, November 2025. Moonshot AI’s most advanced open reasoning model to date, extending the K2 series into agentic, long-horizon reasoning. (262K-token context; API price $0.6 in and $2.5 out per 1M tokens; open weights)
Nova Premier 1.0 — Amazon, October 2025. Amazon Nova Premier is the most capable of Amazon’s multimodal models for complex reasoning tasks and for use as the best teacher for distilling custom models. (1M-token context; API price $2.5 in and $12.5 out per 1M tokens)
Sonar Pro Search — Perplexity, October 2025. Exclusively available on the OpenRouter API, Sonar Pro's new Pro Search mode is Perplexity's most advanced agentic search system. (200K-token context; API price $3 in and $15 out per 1M tokens)
Voxtral Small 24B 2507 — Mistral, October 2025. Enhancement of Mistral Small 3, incorporating state-of-the-art audio input capabilities while retaining best-in-class text performance. (32K-token context; API price $0.1 in and $0.3 out per 1M tokens; open weights)
gpt-oss-safeguard-20b — OpenAI, October 2025. Safety reasoning model from OpenAI built upon gpt-oss-20b. (131K-token context; API price $0.075 in and $0.3 out per 1M tokens; open weights)
MiniMax M2 — MiniMax, October 2025. Compact, high-efficiency large language model optimized for end-to-end coding and agentic workflows. (204K-token context; API price $0.255 in and $1.02 out per 1M tokens; open weights)
Granite 4.0 Micro — IBM, October 2025. 3B parameter from the Granite 4 family of models. (131K-token context; API price $0.017 in and $0.112 out per 1M tokens; open weights)
GPT-5 Image Mini — OpenAI, October 2025. GPT-5 Image Mini combines OpenAI's advanced language capabilities, powered by GPT-5 Mini, with GPT Image 1 Mini for efficient image generation. (400K-token context; API price $2.5 in and $2 out per 1M tokens)
Claude Haiku 4.5 — Anthropic, October 2025. Anthropic's fastest, lowest-cost model, close to frontier quality. (200K-token context; API price $1 in and $5 out per 1M tokens)
GPT-5 Image — OpenAI, October 2025. GPT-5 Image combines OpenAI's GPT-5 model with state-of-the-art image generation capabilities. (400K-token context; API price $10 in and $10 out per 1M tokens)
Qwen3 VL 8B Instruct — Alibaba, October 2025. Multimodal vision-language model from the Qwen3-VL series, built for high-fidelity understanding and reasoning across text, images, and video. (262K-token context; API price $0.25 in and $0.75 out per 1M tokens; open weights)
Llama 4 Maverick — Meta, April 2025. Meta's larger open multimodal model. (524K-token context; API price $0.188 in and $0.652 out per 1M tokens; open weights)
Llama 4 Scout — Meta, April 2025. Meta's smaller open model, with a very long context. (1.3M-token context; API price $0.1 in and $0.3 out per 1M tokens; open weights)