AI Models Compared
Every major frontier model, grouped by provider - pricing, context window, benchmark scores, and what each one is actually best at. Continuously web-verified; last checked 2026-11-17.
Recently released
Pricing at a glance
| Model | Strong at | Context | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|---|
| OpenAI (23 models) | |||||
| GPT-6 CyberNewPreview | - | 1.05M | $20.00 | $100.00 | Closed |
| GPT-Image-2.6 FlareNew | - | Not published | $8.00 | $30.00 | Closed |
| GPT-Image-2.6 SunburstNew | - | Not published | $8.00 | $30.00 | Closed |
| GPT-Live-1.1New | - | Not published | Not announced | Not announced | Closed |
| GPT-6.1 TerraNew | - | 1.05M | $1.00 | $5.00 | Closed |
| GPT-6 RosalindNewPreview | - | Not announced | $15.00 | $75.00 | Closed |
| GPT-6 LunaNew | Cost efficiency | 1.05M | $0.10 | $0.50 | Closed |
| GPT-6.1 SolNew | Long context | 1,050,000 | $2.00 | $10.00 | Closed |
| GPT-6 SolNewLegacy | Cost efficiency | 1.05M | $2.00 | $10.00 | Closed |
| GPT-Image-2.5 FlareNewLegacy | - | Not published | $8.00 | $30.00 | Closed |
| GPT-6 AstraNew | Coding | 1.05M | $10.00 | $50.00 | Closed |
| GPT-Live-1NewLegacy | Speed | Not published | $0.05 per session minute | Included (see note) | Closed |
| GPT-5.6-CyberPreview | Coding | 400K | $12.50 | $75.00 | Closed |
| GPT-Realtime-2.1 | - | 128K | $32.00 | $64.00 | Closed |
| GPT-Transcribe | - | 25 MB audio files | $0.0045 per audio minute | Not announced | Closed |
| GPT-5.6Legacy | Coding | 1.05M | $4.00 | $20.00 | Closed |
| GPT-5.6 SolLegacy | Coding | 1.05M | $4.00 | $20.00 | Closed |
| GPT-5.5 InstantLegacy | Speed | Not published | ChatGPT only (no API) | ChatGPT only (no API) | Closed |
| GPT-5.5Legacy | Coding | 1,050,000 tokens (128,000 max output) | $5.00 | $30.00 | Closed |
| GPT-5.4Legacy | Long context | 1,050,000 tokens (128,000 max output) | $2.50 | $15.00 | Closed |
| GPT-5Legacy | Math | 400,000 tokens (128,000 max output) | $1.25 | $10.00 | Closed |
| o1Legacy | Reasoning | 200,000 tokens (100,000 max output) | $15.00 | $60.00 | Closed |
| GPT-4oLegacy | Speed | 128,000 tokens (16,384 max output) | $2.50 | $10.00 | Closed |
| Google (15 models) | |||||
| Gemini 4 HydrogenNew | - | 1M (shared input + output) | $1.50 | $7.50 | Closed |
| Gemini Decision 1New | - | 262,144 tokens | $0.10 | N/A (no generated text) | Closed |
| Gemini 4 CarbonNew | - | 1M | $2.00 | $10.00 | Closed |
| Gemini 3.8 Flash CyberNewPreview | - | Not announced | Not announced | Not announced | Closed |
| Gemini 4 ArgonNew | Long context | 1M (shared input + output) | $2.00 | $10.00 | Closed |
| Gemini 3.8 Flash TTSNew | - | 8192 tokens | $0.50 | $9.00 | Closed |
| Gemini 3.8 FlashNewLegacy | Coding | 1M | $0.75 | $3.75 | Closed |
| Gemini 3.5 Transcribe | - | Up to 1 hour of audio per request | $2.00 | $12.00 | Closed |
| Gemini 3.7 FlashLegacy | - | 1M | $0.75 | $3.75 | Closed |
| Gemini 3.5 Flash-Lite | - | 1M | $0.30 | $2.50 | Closed |
| Gemini 3.6 FlashLegacy | Speed | 1M | $0.75 | $3.75 | Closed |
| Gemini 3.5 Flash CyberPreviewLegacy | Coding | Not announced | Not announced | Not announced | Closed |
| Gemini Omni Flash | Multimodal | Not published | $1.50 | $12.00 | Closed |
| Nano Banana 2 Lite | Speed | 65,536 input / 4,096 output tokens | $0.25 | $30.00 | Closed |
| Gemma 4 12B | Cost efficiency | 256K tokens | Free | Free | Open |
| Anthropic (14 models) | |||||
| Claude Fable 5.2New | - | 1M | $8.00 | $40.00 | Closed |
| Claude Mythos 5.2NewPreview | - | 1M | $8.00 | $40.00 | Closed |
| Claude Haiku 5.5New | Cost efficiency | 1M | $0.10 | $0.50 | Closed |
| Claude Sonnet 5.5New | Speed | 1M | $2.00 | $10.00 | Closed |
| Claude Opus 5.5New | Coding | 1M | $4.00 | $20.00 | Closed |
| Claude Mythos 5.1NewPreviewLegacy | Coding | 1M | $10.00 | $50.00 | Closed |
| Claude Fable 5.1NewLegacy | - | 1M | $10.00 | $50.00 | Closed |
| Claude Opus 5Legacy | Long context | 1M | $5.00 | $25.00 | Closed |
| Claude Sonnet 5Legacy | Coding | 1M | $2.00 | $10.00 | Closed |
| Claude Fable 5Legacy | Coding | 1M | $10.00 | $50.00 | Closed |
| Claude Opus 4.8Legacy | Coding | 1M | $5.00 | $25.00 | Closed |
| Claude Opus 4.7Legacy | Coding | 1M | $5.00 | $25.00 | Closed |
| Claude Sonnet 4.6Legacy | Long context | 1M | $3.00 | $15.00 | Closed |
| Claude Haiku 4.5Legacy | Speed | 200K | $1.00 | $5.00 | Closed |
| Meta (7 models) | |||||
| Muse EmberNew | - | 1M | $0.10 | $0.50 | Closed |
| Muse Spark OpenNew | Multimodal | 1M | $1.25 | $4.25 | Open |
| Muse WonderNewPreview | Coding | 1M | $4.00 | $20.00 | Closed |
| Muse Spark 1.3New | - | 1M | $1.25 | $4.25 | Closed |
| Muse Glimmer | Cost efficiency | 128K | Free | Free | Open |
| Muse Image | Cost efficiency | Not published | Not applicable | $0.01/image | Closed |
| Llama 4 | Long context | Up to 10M tokens (Scout); ~1M tokens (Maverick) | Free | Free | Open |
| Mistral AI (7 models) | |||||
| Voxtral 3NewPreview | Cost efficiency | 32K (transcription models) | Per audio minute (see note) | Per character for TTS (see note) | Closed |
| Mistral Large 4New | Cost efficiency | 1M | $0.70 | $2.10 | Open |
| Mistral OCR 4.1 | - | Not published | $4.00 per 1,000 pages | $5.00 per 1,000 annotated pages | Closed |
| Robostral NavigatePreview | Cost efficiency | N/A (vision-to-action) | Contact sales | Contact sales | Closed |
| Mistral OCR 4Legacy | Cost efficiency | 16K | $4.00 per 1,000 pages (API) / $2.00 per 1,000 pages (Batch) | $5.00 per 1,000 pages (Document AI annotation tier) | Closed |
| Mistral Small 4 | - | 256K | $0.15 | $0.60 | Open |
| Mistral Large 3Legacy | Reasoning | 256K | $0.50 | $1.50 | Open |
| Google DeepMind (6 models) | |||||
| Gemini 3.8 LiveNew | Speed | 128K | $3.00 | $12.00 | Closed |
| Gemini Robotics ER 2Preview | Multimodal | 128K | $1.00 | $5.00 | Closed |
| Gemini 3.5Legacy | Coding | 1,048,576 tokens (Gemini 3.5 Flash; Pro variant not yet released) | $1.50 | $9.00 | Closed |
| Gemini 2.5 ProLegacy | Long context | 1,048,576 tokens (1M) input; up to 65K output | $1.25 | $10.00 | Closed |
| Gemini 2.5 FlashLegacy | Cost efficiency | 1,048,576 tokens (1M) input; up to 65,535 output | $0.30 | $2.50 | Closed |
| Gemma 3Legacy | Cost efficiency | 128K tokens (32K for the 1B variant) | Free | Free | Open |
| Alibaba (4 models) | |||||
| Qwen4-MaxNew | - | 1M | $1.50 | $4.50 | Closed |
| Qwen 3.8-Flash-Next | Cost efficiency | 262K | Free | Free | Open |
| Qwen 3.8 27B | Coding | 262K | $0.50 | $3.00 | Open |
| Qwen 3.8 MaxLegacy | Long context | 1M | $2.00 | $6.00 | Closed |
| Cohere (4 models) | |||||
| Cohere Parse 5 | Cost efficiency | 8K | $1.50 per 1,000 pages | N/A (billed per page) | Closed |
| North Mini Code | Cost efficiency | 256K tokens | Free | Free | Open |
| Command A+ | - | 128K | $0.30 | $1.50 | Open |
| Command R+Legacy | Long context | 128K tokens | $2.50 | $10.00 | Closed |
| DeepSeek (4 models) | |||||
| DeepSeek V4.1 FlashNew | - | 1M tokens | $0.30 | $1.20 | Open |
| DeepSeek V4 | Long context | 1M tokens | $1.32 | $3.96 | Open |
| DeepSeek V4 FlashLegacy | Long context | 1M tokens | $0.30 | $1.20 | Open |
| DeepSeek V3Legacy | Cost efficiency | 128K tokens | Free | Free | Open |
| Z.ai (Zhipu AI) (4 models) | |||||
| GLM-5.3 | - | 1M | $1.40 | $4.40 | Open |
| GLM-5.3-Flash | Long context | 1M | $0.15 | $0.50 | Open |
| Ox AlphaLegacy | Long context | 1,048,576 (1M tokens) | Free during stealth period (ended) | Free during stealth period (ended) | Closed |
| GLM-5.2Legacy | Long context | 1M | $1.40 | $4.40 | Open |
| Amazon Web Services (2 models) | |||||
| Amazon Nova 2 Lite | - | 1M | $0.30 | $2.50 | Closed |
| Amazon Nova Pro | Multimodal | 300K tokens | $0.80 | $3.20 | Closed |
| Cognition (2 models) | |||||
| SWE-2 FlashNew | - | Not published | $0.10 | $0.50 | Closed |
| SWE-2New | Coding | Not published | $3.00 | $15.00 | Closed |
| Microsoft (2 models) | |||||
| Microsoft-Decision-1NewPreview | Speed | Not published | $0.042 | N/A (no generated text) | Closed |
| MAI-Cyber-1-FlashPreview | Coding | 256000 | Not announced | Not announced | Closed |
| MiniMax (2 models) | |||||
| MiniMax H3-ImageNew | - | Not published | $1.50 | $3.80 | Open |
| MiniMax M3.1 FlashNew | Cost efficiency | 1M | $0.15 | $0.60 | Open |
| Moonshot AI (2 models) | |||||
| Kimi K3.1 ProNew | - | 1M | $2.00 | $10.00 | Open |
| Kimi K3Legacy | Coding | 1M | $3.00 | $15.00 | Open |
| NVIDIA (2 models) | |||||
| Nemotron 3.5 Bolt 74B-A9BNew | - | 256K | Free | Free | Open |
| Nemotron 3.5 Super VLNew | Cost efficiency | 1M | $0.10 | $0.40 | Open |
| xAI (SpaceXAI) (2 models) | |||||
| Grok Voice 3.0New | - | Not disclosed | Not applicable | Not applicable | Closed |
| Grok 4.7New | Coding | 500K | $2.00 | $6.00 | Closed |
| Aleph Alpha (1 model) | |||||
| Kolibri-1New | Cost efficiency | 262K (validated to 1M) | Free | Free | Open |
| Alibaba (Qwen Team) (1 model) | |||||
| Qwen 3 | Cost efficiency | 128K tokens (32K for 0.6B/1.7B/4B dense variants) | Free | Free | Open |
| Baidu (1 model) | |||||
| ERNIE 5.5 Flash PreviewNewPreview | - | 1M | $0.28 | $1.13 | Closed |
| Canva (1 model) | |||||
| Canva CraftNew | Multimodal | Not published | Not announced | Not announced | Closed |
| Cisco (1 model) | |||||
| Antares-1B | Cost efficiency | 128000 | Free | Free | Open |
| Cloudflare (1 model) | |||||
| ClairicNewPreview | Cost efficiency | 65,536 | $0.10 | N/A (no generated text) | Closed |
| Liquid AI (1 model) | |||||
| LFM3New | - | 256K | $0.08 | $0.20 | Open |
| Poolside (1 model) | |||||
| Laguna S 2.1 | Coding | 1M | $0.09 | $0.18 | Open |
| Reflection AI (1 model) | |||||
| BeamNew | Math | 1M | $0.25 | $1.00 | Open |
| Salesforce (1 model) | |||||
| Salesforce KoaNewPreview | Reasoning | Not published | Not announced | Not announced | Closed |
| SpaceXAI (1 model) | |||||
| Grok 4.6Legacy | Cost efficiency | 500K | $2.00 | $6.00 | Closed |
| SpaceXAI (xAI) (1 model) | |||||
| Grok 4.5Legacy | Cost efficiency | 500K | $2.00 | $6.00 | Closed |
| Springboards (1 model) | |||||
| FlintPreview | Speed | Not disclosed | App-only (no API) | App-only (no API) | Closed |
| StepFun (1 model) | |||||
| Step 5 PreviewNewPreview | Cost efficiency | 1M | $1.00 | $2.70 | Open |
| Tencent (1 model) | |||||
| Hy4New | - | 1M | $0.834 | $2.501 | Open |
| Thomson Reuters (1 model) | |||||
| Thomson | Cost efficiency | 262000 | Not sold via public API | Not sold via public API | Closed |
| TypeSafe AI (1 model) | |||||
| JevNew | Cost efficiency | 32K | $0.042 | Free | Closed |
| Writer (1 model) | |||||
| Palmyra X6 | Cost efficiency | 1000000 | $2.00 | $8.00 | Closed |
Prices in USD. Open-source models are free to self-host; API pricing varies by provider.
OpenAI23 models
GPT-6 CyberNew
1.05M ctx2026-11
OpenAI's cybersecurity variant of GPT-6 Astra, limited to approved Daybreak Red organizations
GPT-Image-2.6 FlareNew
Not published ctx2026-11
OpenAI's fast GPT Image tier for quick iteration, now with GIF and SVG output
GPT-Image-2.6 SunburstNew
Not published ctx2026-11
OpenAI's higher-fidelity GPT Image tier for detailed compositions and editing
GPT-Live-1.1New
Not published ctx2026-10
OpenAI's updated full-duplex voice model for ChatGPT, with better accents and 13 new voices
GPT-6.1 TerraNew
1.05M ctx2026-10
OpenAI's fast, low-cost GPT-6 model for computer use, visual reasoning and high-volume work
GPT-6 RosalindNew
Not announced ctx2026-10
OpenAI's GPT-6 life sciences model, limited to approved research organizations
GPT-6 LunaNew
1.05M ctx2026-09
OpenAI's cheapest GPT-6 model, and since October 2026 the default for free ChatGPT users
GPT-6.1 SolNew
1,050,000 ctx2026-09
Near-Astra intelligence for a fifth of the price
GPT-6 SolNew
1.05M ctx2026-09
OpenAI's September 2026 Sol model for coding and agentic work, since replaced by GPT-6.1 Sol
GPT-Image-2.5 FlareNew
Not published ctx2026-09
OpenAI's September 2026 fast image model, since followed by GPT-Image-2.6 Flare at the same price
GPT-6 AstraNew
1.05M ctx2026-09
OpenAI's most capable model for the most demanding work
GPT-Live-1New
Not published ctx2026-09
A new generation of full-duplex voice models for natural human-AI conversation
GPT-5.6-Cyber
400K ctx2026-08
Purpose-trained model for advanced cybersecurity research and vulnerability testing
GPT-Realtime-2.1
128K ctx2026-07
OpenAI's current speech-to-speech reasoning model for voice agents on the Realtime API
GPT-Transcribe
25 MB audio files ctx2026-07
OpenAI's recommended speech-to-text model for recorded audio and batch transcription
GPT-5.6
1.05M ctx2026-07
OpenAI's GPT-5.6 model family (Sol, Terra, Luna), since followed by GPT-6
GPT-5.6 Sol
1.05M ctx2026-06
OpenAI's GPT-5.6 flagship tier, now generally available and since replaced by GPT-6 Sol
GPT-5.5 Instant
Not published ctx2026-05
OpenAI's fast ChatGPT model from May 2026, no longer the ChatGPT default
GPT-5.5
1,050,000 tokens (128,000 max output) ctx2026-04
OpenAI's April 2026 frontier model for professional work, since replaced by GPT-5.6 Sol and the GPT-6 family
GPT-5.4
1,050,000 tokens (128,000 max output) ctx2026-03
OpenAI's March 2026 model for coding and professional work, since superseded by GPT-5.5 and GPT-6
GPT-5
400,000 tokens (128,000 max output) ctx2025-08
OpenAI's landmark August 2025 flagship: strong reasoning at a low price
o1
200,000 tokens (100,000 max output) ctx2024-12
OpenAI's first-generation deep-reasoning model that thinks before answering
GPT-4o
128,000 tokens (16,384 max output) ctx2024-05
OpenAI's versatile, fast multimodal workhorse (text + image)
Google15 models
Gemini 4 HydrogenNew
1M (shared input + output) ctx2026-11
Google's fast Gemini 4 model for everyday coding and agent work, with a free API tier
Gemini Decision 1New
262,144 tokens ctx2026-11
A Google decision model that returns calibrated probabilities for yes/no, choice and rubric questions instead of text
Gemini 4 CarbonNew
1M ctx2026-10
Google's most capable Gemini model, generally available since November 10, 2026
Gemini 3.8 Flash CyberNew
Not announced ctx2026-09
Google's restricted Flash model for finding and patching software vulnerabilities
Gemini 4 ArgonNew
1M (shared input + output) ctx2026-09
Google's frontier model for real-world coding, enterprise knowledge work, and cybersecurity defense
Gemini 3.8 Flash TTSNew
8192 tokens ctx2026-09
Advanced text-to-speech with natural language voice design and line-by-line performance direction
Gemini 3.8 FlashNew
1M ctx2026-09
Google's previous-generation Flash model for coding and agents, at half Gemini 4 Hydrogen's price until December 31, 2026
Gemini 3.5 Transcribe
Up to 1 hour of audio per request ctx2026-08
Google's speech-to-text model that turns raw audio into formatted text
Gemini 3.7 Flash
1M ctx2026-08
Most intelligent workhorse model yet for coding and agents
Gemini 3.5 Flash-Lite
1M ctx2026-07
Google's fastest, lowest-cost Gemini 3.5-class model for high-volume agent and document work
Gemini 3.6 Flash
1M ctx2026-07
Google's previous-generation Flash model, still served in the Gemini API
Gemini 3.5 Flash Cyber
Not announced ctx2026-07
Cost-efficient cybersecurity-focused model for vulnerability detection and patching
Gemini Omni Flash
Not published ctx2026-06
Create anything from any input with conversational video editing
Nano Banana 2 Lite
65,536 input / 4,096 output tokens ctx2026-06
The fastest, most cost-efficient image generation model in the Nano Banana family
Gemma 4 12B
256K tokens ctx2026-06
OpenGoogle's laptop-runnable open multimodal model with a unified encoder-free design.
Anthropic14 models
Claude Fable 5.2New
1M ctx2026-11
Anthropic's top Claude tier for long-horizon agentic work, 20% cheaper than Fable 5.1 at $8/$40
Claude Mythos 5.2New
1M ctx2026-11
Claude Fable 5.2 with fewer safeguards, for organizations in Anthropic's verification programs
Claude Haiku 5.5New
1M ctx2026-10
Anthropic's fastest, lowest-cost current model at $0.10 per million input tokens with a 1M context window
Claude Sonnet 5.5New
1M ctx2026-09
Faster, cheaper mid-tier model for everyday tasks and agentic work
Claude Opus 5.5New
1M ctx2026-09
Anthropic's current Opus and its recommended starting model, priced at $4/$20 per 1M tokens
Claude Mythos 5.1New
1M ctx2026-09
Claude Fable 5.1 with relaxed safeguards for verified organizations, replaced by Mythos 5.2 in November 2026
Claude Fable 5.1New
1M ctx2026-09
Anthropic's September 2026 Fable model, replaced by Fable 5.2 on November 9, 2026 and still in the API at $10/$50
Claude Opus 5
1M ctx2026-07
Anthropic's July 2026 Opus for reasoning and agentic work, now replaced by the cheaper Opus 5.5.
Claude Sonnet 5
1M ctx2026-06
Anthropic's June 2026 Sonnet with near-Opus quality, now replaced by Sonnet 5.5 at the same price.
Claude Fable 5
1M ctx2026-06
Anthropic's June 2026 Fable model, now legacy; Anthropic recommends migrating to Fable 5.2 ($8/$40)
Claude Opus 4.8
1M ctx2026-05
Anthropic's May 2026 Opus for agentic coding and enterprise work, now a legacy model replaced by Opus 5 and Opus 5.5.
Claude Opus 4.7
1M ctx2026-04
The April 2026 Opus with improved high-resolution vision, now a legacy model; Opus 5.5 is the current Opus.
Claude Sonnet 4.6
1M ctx2026-02
Anthropic's February 2026 Sonnet, now a legacy model; Sonnet 5.5 is the current Sonnet.
Claude Haiku 4.5
200K ctx2025-10
Anthropic's October 2025 Haiku, now a legacy model replaced by the cheaper Haiku 5.5.
Meta7 models
Muse EmberNew
1M ctx2026-11
Meta's smallest, cheapest Muse model, built for the repeated steps inside agents
Muse Spark OpenNew
1M ctx2026-10
OpenMeta's open-weight release of Muse Spark: an 816B multimodal MoE
Muse WonderNew
1M ctx2026-10
Meta's first frontier model, built for long-running agent tasks
Muse Spark 1.3New
1M ctx2026-09
Meta's proprietary agentic coding model for Muse Code and the Meta Model API
Muse Glimmer
128K ctx2026-08
OpenOpen-weight agentic AI that runs on consumer GPUs and laptops
Muse Image
Not published ctx2026-07
Image generation model with visual reasoning capabilities built for Meta's ecosystem
Llama 4
Up to 10M tokens (Scout); ~1M tokens (Maverick) ctx2025-04
OpenMeta's natively multimodal open-weight MoE models, with up to 10M tokens of context on Scout.
Mistral AI7 models
Voxtral 3New
32K (transcription models) ctx2026-10
Mistral's speech-to-text, text-to-speech and voice agent model family
Mistral Large 4New
1M ctx2026-10
OpenMistral's 1-trillion-parameter open-weight flagship, nicknamed 'le Chonk'
Mistral OCR 4.1
Not published ctx2026-07
Mistral's current OCR model, adding paragraph-level bounding boxes and block confidence scores
Robostral Navigate
N/A (vision-to-action) ctx2026-07
8B embodied navigation model for autonomous robot movement using single RGB camera and natural language instructions
Mistral OCR 4
16K ctx2026-06
Mistral's June 2026 document OCR model, retired in September 2026 and replaced by OCR 4.1
Mistral Small 4
256K ctx2026-03
OpenMistral's open-weight 119B MoE that handles chat, reasoning, image input and coding in one model
Mistral Large 3
256K ctx2025-12
OpenMistral Large 3, the previous open-weight Mistral flagship, now deprecated in favor of Mistral Large 4
Google DeepMind6 models
Gemini 3.8 LiveNew
128K ctx2026-09
Most advanced live dialogue model for natural, real-time voice conversations
Gemini Robotics ER 2
128K ctx2026-07
High-level reasoning brain for robots enabling video understanding, task orchestration, and multi-robot collaboration
Gemini 3.5
1,048,576 tokens (Gemini 3.5 Flash; Pro variant not yet released) ctx2026-05
Google's frontier model for agents and coding, made fast and cheap.
Gemini 2.5 Pro
1,048,576 tokens (1M) input; up to 65K output ctx2025-06
Google's advanced thinking model for complex reasoning, coding, and long context.
Gemini 2.5 Flash
1,048,576 tokens (1M) input; up to 65,535 output ctx2025-06
Google's price-performance workhorse with thinking and a 1M-token context.
Gemma 3
128K tokens (32K for the 1B variant) ctx2025-03
OpenGoogle's previous-generation open, multimodal, multilingual model family.
Alibaba4 models
Qwen4-MaxNew
1M ctx2026-11
Alibaba's API-only Qwen4 flagship with image and video input and a 1M-token context
Qwen 3.8-Flash-Next
262K ctx2026-08
OpenA 125B MoE architecture preview of Qwen4 with only 6B active parameters per token
Qwen 3.8 27B
262K ctx2026-08
OpenApache 2.0 dense multimodal model designed for local deployment on consumer hardware
Qwen 3.8 Max
1M ctx2026-08
Alibaba's 2.4T-parameter multimodal flagship with a 1M-token context and open core weights
Cohere4 models
Cohere Parse 5
8K ctx2026-08
Cost-effective document parsing at enterprise scale
North Mini Code
256K tokens ctx2026-06
OpenCohere's first open-weight agentic coding model - 30B MoE, 3B active, runs on one H100.
Command A+
128K ctx2026-05
OpenCohere's first mixture-of-experts Command model: open weights, vision input and reasoning on two H100s
Command R+
128K tokens ctx2024-08
Cohere's 2024 RAG and tool-use model, still live on the Cohere API but superseded by Command A and Command A+
DeepSeek4 models
DeepSeek V4.1 FlashNew
1M tokens ctx2026-09
OpenDeepSeek's multimodal Flash model with a 1M context, MIT weights and very low API prices
DeepSeek V4
1M tokens ctx2026-08
OpenOpen-weight 1.6T-param MoE frontier model with a 1M-token context built for agents.
DeepSeek V4 Flash
1M tokens ctx2026-07
OpenDeepSeek's 284B open-weight Flash model, replaced by V4.1 Flash in September 2026
DeepSeek V3
128K tokens ctx2024-12
OpenThe open-weight 671B-param MoE that put DeepSeek on the frontier map.
Z.ai (Zhipu AI)4 models
GLM-5.3
1M ctx2026-08
OpenZ.ai's post-trained successor to GLM-5.2, much stronger at long-horizon coding and cyber tasks
GLM-5.3-Flash
1M ctx2026-08
OpenCost-optimized mixture-of-experts model for multimodal coding and long-context tasks
Ox Alpha
1,048,576 (1M tokens) ctx2026-08
Former anonymous OpenRouter stealth model with 1M context, revealed as Z.ai GLM-5.3 Flash
GLM-5.2
1M ctx2026-06
OpenZ.ai's MIT-licensed 1M-context coding model, succeeded by GLM-5.3 in August 2026
Amazon Web Services2 models
Amazon Nova 2 Lite
1M ctx2025-12
Amazon's low-cost reasoning model on Bedrock, with a 1M-token context and built-in tools
Amazon Nova Pro
300K tokens ctx2024-12
Amazon's balanced multimodal Bedrock model for text, image, and video at scale.
Cognition2 models
SWE-2 FlashNew
Not published ctx2026-11
Cognition's fast, low-cost coding model for Devin, at one-fifth of SWE-2's price
SWE-2New
Not published ctx2026-09
Frontier coding performance at 64% lower cost
Microsoft2 models
Microsoft-Decision-1New
Not published ctx2026-10
A decision model that returns calibrated probabilities instead of text, for routing and classification in agent pipelines
MAI-Cyber-1-Flash
256000 ctx2026-07
Microsoft's first dedicated cybersecurity AI model combining efficiency and performance
MiniMax2 models
MiniMax H3-ImageNew
Not published ctx2026-11
OpenMiniMax's open-weight image model for 4K generation and editing in under 10 seconds
MiniMax M3.1 FlashNew
1M ctx2026-10
OpenMiniMax's cheap 1M-context multimodal model for agents and coding
Moonshot AI2 models
Kimi K3.1 ProNew
1M ctx2026-10
OpenMoonshot AI's open-weight 1M-context flagship for long-horizon coding
Kimi K3
1M ctx2026-07
OpenMoonshot AI's 2.8T-parameter open-weight model with a 1M context, succeeded by Kimi K3.1 Pro
NVIDIA2 models
Nemotron 3.5 Bolt 74B-A9BNew
256K ctx2026-10
OpenNVIDIA's open 74B multimodal agent model, with a 2-bit build that fits in about 24GB
Nemotron 3.5 Super VLNew
1M ctx2026-10
OpenNVIDIA's open 120B multimodal reasoning model for agents, sized for a single GPU
xAI (SpaceXAI)2 models
Grok Voice 3.0New
Not disclosed ctx2026-10
xAI's speech-to-speech model for voice agents, ranked #1 on Artificial Analysis's speech index at launch
Grok 4.7New
500K ctx2026-09
SpaceXAI's most capable model for coding and knowledge work
Aleph Alpha1 model
Alibaba (Qwen Team)1 model
Baidu1 model
Canva1 model
Cisco1 model
Cloudflare1 model
Liquid AI1 model
Poolside1 model
Reflection AI1 model
Salesforce1 model
SpaceXAI1 model
SpaceXAI (xAI)1 model
Springboards1 model
StepFun1 model
Tencent1 model
Thomson Reuters1 model
TypeSafe AI1 model
Writer1 model
Benchmark scores sourced from official provider release posts. Prices subject to change - check provider pricing pages for current rates.