Models and prices · prices checked 19 Oct, 23:40 UTC
What each model costs.
Every current model from the 13 ranked labs, with list prices per million tokens. Prices come from each lab's own pricing page where we can read it, otherwise from the lab's own endpoint; the linked pricing page is always the source of truth. DeepSeek shows its off-peak price, which applies most of the day.
Quality against price
Each model's rating on Arena's text leaderboard against its blended price (three parts input to one part output, per million tokens). The blue line is the frontier: no cheaper model rates higher. Hover or tab through the points for each model.
Show as a table
| Model | Lab | Arena | Blended price | Frontier |
|---|---|---|---|---|
| Gemini 4 Argon | 1525 | $4 | Yes | |
| Claude Opus 5.5 | Anthropic | 1507 | $8 | |
| Claude Opus 4.6 | Anthropic | 1504 | $10 | |
| Claude Fable 5 | Anthropic | 1504 | $20 | |
| Claude Opus 4.7 | Anthropic | 1501 | $10 | |
| Claude Fable 5.1 | Anthropic | 1501 | $20 | |
| Gemini 3.8 Flash | 1497 | $1.5 | Yes | |
| Muse Spark 1.3 | Meta | 1494 | $2 | |
| Muse Spark 1.2 | Meta | 1492 | $2 | |
| Muse Spark 1.1 | Meta | 1491 | $2 | |
| Claude Opus 5 | Anthropic | 1490 | $10 | |
| Gemini 3.1 Pro Preview | 1487 | $2.3 | ||
| Gemini 3.7 Flash | 1486 | $1.5 | ||
| Gemini 3 Pro | 1485 | $4.5 | ||
| GPT-5.6 Sol | OpenAI | 1485 | $8 | |
| GPT-6.1 Sol | OpenAI | 1484 | $4 | |
| Qwen3.8 | Alibaba | 1483 | $3 | |
| Gemini 3.6 Flash | 1482 | $1.5 | ||
| GPT-5.5 | OpenAI | 1482 | $11 | |
| Claude Opus 4.8 | Anthropic | 1481 | $10 | |
| GLM 5.3 | Z.ai | 1478 | $2.1 | |
| Gemini 3.5 Flash | 1478 | $1.7 | ||
| GPT-5.2 Chat Latest | OpenAI | 1476 | $4.8 | |
| Claude Sonnet 5.5 | Anthropic | 1476 | $4 | |
| GLM-5.2 | Z.ai | 1475 | $2.1 | |
| GPT-5.4 | OpenAI | 1475 | $5.6 | |
| Qwen3.7 Max Preview | Alibaba | 1475 | $2.2 | |
| GLM 5.3 Flash | Z.ai | 1475 | $0.1 | Yes |
| DeepSeek V4.1 Flash | DeepSeek | 1475 | $0.52 | |
| Claude Opus 4.5 20251101 High 32K | Anthropic | 1474 | $10 | |
| Gemini 3 Flash | 1473 | $1.1 | ||
| Claude Sonnet 4.6 | Anthropic | 1472 | $3 | |
| Grok 4.20 Beta | xAI | 1472 | $1.6 | |
| Grok 4.20 Multi Agent Beta | xAI | 1470 | $1.6 | |
| Claude Opus 4.5 | Anthropic | 1470 | $10 | |
| GPT-5.6 Terra | OpenAI | 1466 | $4.5 | |
| Grok 4.5 | xAI | 1466 | $3 | |
| Qwen3.5 Max Preview | Alibaba | 1465 | $2.4 | |
| DeepSeek-V4-Pro | DeepSeek | 1465 | $2.0 | |
| GLM-5.1 | Z.ai | 1464 | $2.1 | |
| Claude Sonnet 5 | Anthropic | 1461 | $4 | |
| Kimi K2.6 | Moonshot | 1461 | $1.7 | |
| Qwen3.6 Max Preview | Alibaba | 1460 | $2.9 | |
| DeepSeek-V4-Pro | DeepSeek | 1458 | $2.0 | |
| GLM-5 | Z.ai | 1458 | $1.6 | |
| Claude Sonnet 4.5 20250929 High 32K | Anthropic | 1456 | $6 | |
| Qwen3.7 Plus | Alibaba | 1456 | $0.56 | |
| GPT-6 Sol | OpenAI | 1456 | $4 | |
| GPT-5.1 | OpenAI | 1456 | $1.7 | |
| Claude Sonnet 4.5 | Anthropic | 1455 | $6 | |
| Gemini 3.5 Flash-Lite | 1455 | $0.42 |
| Model | Lab | Released | Input | Cached input | Output | Context | Recent change |
|---|---|---|---|---|---|---|---|
| GPT-6 LunaHigh-volume apps, extraction and classification at the lowest GPT-6 price.Docs · Pricing · Best practices | OpenAI | September 2026 | $0.10 | $0.01 | $0.50 | 1.05M | |
| GPT-6 SolEveryday production work that needs strong quality at a moderate price.Docs · Pricing · Best practices | OpenAI | September 2026 | $2.00 | $0.20 | $10.00 | 1.05M | |
| GPT-6 AstraThe hardest reasoning, coding and long agentic tasks.Docs · Pricing · Best practices | OpenAI | September 2026 | $10.00 | $1.00 | $50.00 | 1.05M | |
| GPT-5.6 SolTeams tackling the hardest agentic and coding work.Docs · Pricing · Best practices | OpenAI | Jul 2026 | $4.00 | $0.40 | $20.00 | 1M | |
| GPT-5.6 TerraEveryday production apps that need strong quality on a budget.Docs · Pricing · Best practices | OpenAI | Jul 2026 | $2.50 | · | $15.00 | 1M | |
| GPT-5.6 LunaChat, classification, and other high-throughput jobs.Docs · Pricing · Best practices | OpenAI | Jul 2026 | $1.00 | · | $6.00 | 1M | |
| GPT-5.5Broad production use across chat and tools.Docs · Pricing · Best practices | OpenAI | Apr 2026 | $5.00 | · | $30.00 | 1M | |
| GPT-5.5 ProResearch-grade analysis where accuracy beats cost.Docs · Pricing · Best practices | OpenAI | Apr 2026 | $30.00 | · | $180.00 | 1M | |
| GPT-5.4 miniLightweight tasks and cost-sensitive apps.Docs · Pricing · Best practices | OpenAI | March 2026 | $0.75 | · | $4.50 | 1M | |
| GPT-5.4 nanoClassification, extraction, and routing at scale.Docs · Pricing · Best practices | OpenAI | March 2026 | $0.20 | · | $1.25 | 1M | |
| o3Math, logic, and analysis that need deeper thinking.Docs · Pricing · Best practices | OpenAI | Apr 2025 | $2.00 | · | $8.00 | 200K | |
| GPT-Realtime-2Voice agents and live conversation. Prices are per 1M audio tokens.Docs · Pricing · Best practices | OpenAI | $32.00 | · | $64.00 | 32K | ||
| GPT-image-2Product imagery, design, and creative work. Text in, image tokens out.Docs · Pricing · Best practices | OpenAI | $5.00 | · | $30.00 | N/A | ||
| gpt-oss-120b Open weightsTeams that want to self-host a capable model.Docs · Pricing · Best practices | OpenAI | Aug 2025 | Open | · | Open | 128K | |
| gpt-oss-20b Open weightsOn-device and low-footprint deployments.Docs · Pricing · Best practices | OpenAI | Aug 2025 | Open | · | Open | 128K | |
| Gemini 4 ArgonDocs · Pricing | October 2026 | $2.00 | $0.10 | $10.00 | 1M | ||
| Nano Banana 2.1Docs · Pricing | October 2026 | $1.50 | · | $7.50 | 64K | ||
| Gemini 3.8 FlashFast, affordable agentic and coding work.Docs · Pricing · Best practices | September 2026 | $0.75 | $0.075 | $3.75 | 1M | ||
| Gemini 3.7 FlashFast multimodal work where 3.8 Flash is not yet needed.Docs · Pricing · Best practices | August 2026 | $0.75 | $0.075 | $3.75 | 1M | ||
| Gemini 3.1 ProComplex analysis over very long documents. Rates rise above 200K tokens.Docs · Pricing · Best practices | $2.00 | · | $12.00 | 2M | |||
| Gemini 3.6 FlashFast, affordable agentic and coding work.Docs · Pricing · Best practices | Jul 2026 | $0.75 | $0.075 | $3.75 | 1M | ||
| Gemini 3.5 FlashGeneral production workloads.Docs · Pricing · Best practices | May 2026 | $1.50 | $0.15 | $9.00 | 1M | ||
| Gemini 3.5 Flash-LiteLarge-scale, latency-sensitive tasks.Docs · Pricing · Best practices | Jul 2026 | $0.30 | $0.03 | $2.50 | 1M | ||
| Gemini 3.5 Flash CyberSecurity teams. No public API pricing or self-serve access.Docs · Pricing · Best practices | Jul 2026 | See pricing | · | See pricing | 1M | ||
| Claude Sonnet 5.5Docs · Pricing | Anthropic | September 2026 | $2.00 | $0.20 | $10.00 | 1M | |
| Claude Opus 5.5Agentic coding and enterprise work at a lower price than Opus 5.Docs · Pricing · Best practices | Anthropic | September 2026 | $4.00 | $0.20 | $20.00 | 1M | |
| Claude Fable 5.1Long-running agents and the hardest knowledge work.Docs · Pricing · Best practices | Anthropic | September 2026 | $10.00 | $0.25 | $50.00 | 1M | |
| Claude Opus 5Teams already built on Opus 5.Docs · Pricing · Best practices | Anthropic | July 2026 | $5.00 | $0.50 | $25.00 | 1M | |
| Claude Fable 5Ambitious, long-horizon agent work.Docs · Pricing · Best practices | Anthropic | Jun 2026 | $10.00 | $1.00 | $50.00 | 1M | |
| Claude Mythos 5.1Approved security teams in Project Glasswing. Invitation only.Docs · Pricing · Best practices | Anthropic | September 2026 | $10.00 | $0.25 | $50.00 | 1M | |
| Claude Mythos 5Approved security teams in Project Glasswing. Invitation only.Docs · Pricing · Best practices | Anthropic | Jun 2026 | $10.00 | $1.00 | $50.00 | 1M | |
| Claude Opus 4.8Complex agentic coding and enterprise work.Docs · Pricing · Best practices | Anthropic | May 2026 | $5.00 | $0.50 | $25.00 | 1M | |
| Claude Sonnet 5Most production workloads.Docs · Pricing · Best practices | Anthropic | June 2026 | $2.00 | $0.20 | $10.00 | 1M | |
| Claude Haiku 4.5Latency-sensitive and high-volume tasks.Docs · Pricing · Best practices | Anthropic | Oct 2025 | $1.00 | $0.10 | $5.00 | 200K | |
| Grok 4.7Coding and agent work with a large context window.Docs · Pricing · Best practices | xAI | September 2026 | $2.00 | $0.50 | $6.00 | 500K | ▲ 25%was $1.60 / $4.80 to 28 Sep |
| Grok 4.6Teams already built on Grok 4.6.Docs · Pricing | xAI | August 2026 | $2.00 | $0.50 | $6.00 | 500K | |
| Grok 4.5Hard coding and agent tasks. Cached input is $0.50 per 1M.Docs · Pricing | xAI | Jul 2026 | $2.00 | $0.30 | $6.00 | 500K | |
| Grok 4.3General production use. Canonical chat-model alias on the docs.Docs · Pricing | xAI | Apr 2026 | $1.25 | $0.20 | $2.50 | 1M | |
| Grok 4.20Multi-agent and long-horizon jobs.Docs · Pricing | xAI | March 2026 | $1.25 | $0.20 | $2.50 | 2M | |
| Grok 4.1 FastHigh-volume, cost-sensitive tasks.Docs · Pricing | xAI | $0.20 | · | $0.50 | 2M | ||
| Grok Build 0.1Codebase-scale software work.Docs · Pricing | xAI | May 2026 | $1.00 | $0.20 | $2.00 | 256K | |
| Grok ImagineCreative and media work.Docs · Pricing · Best practices | xAI | See pricing | · | See pricing | N/A | ||
| Muse Spark Open Open weightsDocs · Pricing | Meta | October 2026 | Open | · | Open | 1M | |
| Muse WonderDocs · Pricing | Meta | October 2026 | $4.00 | $0.20 | $20.00 | 1M | |
| Muse Spark 1.3Agentic and coding work on Meta's API.Docs · Pricing · Best practices | Meta | September 2026 | $1.25 | $0.15 | $4.25 | 1M | |
| Muse Glimmer 30B Open weightsSelf-hosted agents that need an open multimodal model.Docs · Pricing · Best practices | Meta | August 2026 | Open | · | Open | 128K | |
| Muse Spark 1.2Teams already built on Muse Spark 1.2.Docs · Pricing · Best practices | Meta | August 2026 | $1.25 | $0.15 | $4.25 | 1M | |
| Llama 4 Maverick Open weightsGeneral assistant and multimodal workloads run through third-party hosts or self-hosting.Docs · Pricing · Best practices | Meta | April 2025 | Open | · | Open | 1M | |
| Llama 4 Scout Open weightsVery long context tasks and efficient on device or single GPU deployment.Docs · Pricing · Best practices | Meta | April 2025 | Open | · | Open | 10M | |
| Llama 3.3 70B Open weightsTeams wanting a proven mid size open text model.Docs · Pricing · Best practices | Meta | December 2024 | Open | · | Open | 128K | |
| Llama 3.2 90B Vision Open weightsMultimodal apps that need self hosted image plus text.Docs · Pricing · Best practices | Meta | September 2024 | Open | · | Open | 128K | |
| Llama 3.2 11B Vision Open weightsLighter multimodal workloads and edge image tasks.Docs · Pricing · Best practices | Meta | September 2024 | Open | · | Open | 128K | |
| Llama 3.2 3B Open weightsEdge and mobile assistants where footprint matters.Docs · Pricing · Best practices | Meta | September 2024 | Open | · | Open | 128K | |
| Llama 3.1 405B Open weightsResearch and high end self hosted deployments.Docs · Pricing · Best practices | Meta | July 2024 | Open | · | Open | 128K | |
| Llama 3.1 8B Open weightsCost sensitive text tasks and fine tuning starters.Docs · Pricing · Best practices | Meta | July 2024 | Open | · | Open | 128K | |
| Mistral Large 4Docs · Pricing | Mistral | October 2026 | $0.68 | $0.07 | $2.09 | 512K | |
| Mistral Medium 3.5Complex agent pipelines and premium coding assistants.Docs · Pricing · Best practices | Mistral | April 2026 | $1.50 | · | $7.50 | 256K | |
| Mistral Large 3 Open weightsTeams wanting a top open weight generalist with a hosted option.Docs · Pricing · Best practices | Mistral | December 2025 | $0.50 | · | $1.50 | 256K | |
| Mistral Small 4 Open weightsHigh throughput budget tasks that still need long context.Docs · Pricing · Best practices | Mistral | December 2025 | $0.15 | $0.015 | $0.60 | 256K | |
| CodestralIDE autocomplete and high frequency code generation.Docs · Pricing · Best practices | Mistral | August 2025 | $0.30 | $0.03 | $0.90 | 256K | |
| Devstral MediumAutonomous coding agents and repo scale edits.Docs · Pricing · Best practices | Mistral | 2025 | $0.40 | · | $2.00 | 256K | |
| Devstral Small Open weightsSelf hosted coding agents on a budget.Docs · Pricing · Best practices | Mistral | 2025 | $0.10 | · | $0.30 | 256K | |
| Ministral 3 14B Open weightsOn device and latency sensitive text tasks.Docs · Pricing · Best practices | Mistral | December 2025 | Open | · | Open | 128K | |
| Pixtral 12B Open weightsImage plus text apps that want open weights.Docs · Pricing · Best practices | Mistral | September 2024 | $0.15 | · | $0.15 | 128K | |
| Voxtral Mini Open weightsVoice interfaces and transcription pipelines.Docs · Pricing · Best practices | Mistral | July 2025 | Open | · | Open | 32K | |
| Mistral Large 2 Open weightsWorkloads pinned to the older Large line.Docs · Pricing · Best practices | Mistral | November 2024 | $2.00 | · | $6.00 | 128K | |
| DeepSeek V4.1 Flash 1017 Open weightsOff-peak price, which applies most of the day. Peak hours (01:00 to 04:00 and 06:00 to 10:00 UTC, weekdays) cost double: $0.30 / $1.20.Docs · Pricing | DeepSeek | October 2026 | $0.15 | $0.003 | $0.60 | 1M | |
| DeepSeek V4.1 Flash Open weightsVery low-cost general work; half price off-peak.Off-peak price, which applies most of the day. Peak hours (01:00 to 04:00 and 06:00 to 10:00 UTC, weekdays) cost double: $0.30 / $1.20.Docs · Pricing · Best practices | DeepSeek | September 2026 | $0.15 | $0.003 | $0.60 | 1M | |
| DeepSeek-R1 Open weightsSelf hosted reasoning and research on open weights.Docs · Pricing · Best practices | DeepSeek | January 2025 | Open | · | Open | 128K | |
| DeepSeek-V3.1 Open weightsSelf hosting the earlier V3 line.Docs · Pricing · Best practices | DeepSeek | August 2025 | Open | · | Open | 128K | |
| Qwen3.8 Max PrimeDocs · Pricing | Alibaba | September 2026 | $4.00 | $0.50 | $12.00 | 1M | |
| Qwen3.8 Omni FlashVoice, audio and video agents at a low price.Docs · Pricing · Best practices | Alibaba | September 2026 | $0.15 | $0.016 | $0.47 | 1M | |
| Qwen3.8 MaxThe hardest reasoning and coding on Alibaba Cloud.Docs · Pricing · Best practices | Alibaba | September 2026 | $2.00 | $0.25 | $6.00 | 1M | |
| Qwen3.8 Flash Open weightsHigh-volume multimodal work.Docs · Pricing · Best practices | Alibaba | August 2026 | $0.15 | $0.016 | $0.47 | 1M | |
| Qwen3.8 27B Open weightsSelf-hosting a capable multimodal model on modest hardware.Docs · Pricing · Best practices | Alibaba | August 2026 | $0.42 | $0.085 | $2.55 | 1M | |
| Qwen3.8 2.4T A95B Open weightsTeams that want Max-class quality they can host themselves.Docs · Pricing · Best practices | Alibaba | August 2026 | $2.00 | $0.25 | $6.00 | 1M | |
| Qwen3.7 FlashThe cheapest capable multimodal calls.Docs · Pricing · Best practices | Alibaba | July 2026 | $0.03 | $0.006 | $0.13 | 1M | |
| Qwen3-MaxThe hardest reasoning and long context workloads on Model Studio.Docs · Pricing · Best practices | Alibaba | September 2025 | $0.78 | $0.16 | $3.90 | 256K | |
| Qwen-PlusGeneral purpose production traffic.Docs · Pricing · Best practices | Alibaba | September 2025 | $0.26 | · | $0.78 | 1M | |
| Qwen-FlashSimple tasks that need fast cheap responses.Docs · Pricing · Best practices | Alibaba | 2025 | $0.05 | · | $0.40 | 1M | |
| Qwen-TurboExisting integrations that have not yet migrated to Flash.Docs · Pricing · Best practices | Alibaba | See pricing | · | See pricing | 1M | ||
| Qwen3-Coder-480B-A35B Open weightsAgentic coding, tool use and browser automation.Docs · Pricing · Best practices | Alibaba | July 2025 | Open | · | Open | 256K | |
| Qwen3-Coder-Next Open weightsLighter self hosted coding agents.Docs · Pricing · Best practices | Alibaba | 2026 | Open | · | Open | 256K | |
| Qwen3-VL Open weightsMultimodal document and UI understanding.Docs · Pricing · Best practices | Alibaba | 2025 | Open | · | Open | 256K | |
| Qwen3.6-27B Open weightsStrong coding on a single accessible open model.Docs · Pricing · Best practices | Alibaba | 2026 | Open | · | Open | 256K | |
| Qwen3-235B-A22B Open weightsSelf hosted general reasoning at scale.Docs · Pricing · Best practices | Alibaba | April 2025 | Open | · | Open | 256K | |
| Qwen2.5-MaxIntegrations pinned to the Qwen2.5 line.Docs · Pricing · Best practices | Alibaba | January 2025 | See pricing | · | See pricing | 32K | |
| Command A+ Open weightsEnterprises that need top quality with self hosting options.Docs · Pricing · Best practices | Cohere | May 2026 | $0.30 | $0.15 | $1.50 | 128K | |
| Command A Open weightsEnterprise RAG, agents and general chat.Docs · Pricing · Best practices | Cohere | March 2025 | $2.50 | · | $10.00 | 256K | |
| Command A ReasoningAgentic workflows that need deliberate reasoning.Docs · Pricing · Best practices | Cohere | August 2025 | See pricing | · | See pricing | 256K | |
| Command A VisionEnterprise multimodal document analysis.Docs · Pricing · Best practices | Cohere | July 2025 | See pricing | · | See pricing | 128K | |
| Command R+ Open weightsExisting RAG stacks built on the R+ line.Docs · Pricing · Best practices | Cohere | August 2024 | $2.50 | · | $10.00 | 128K | |
| Command R Open weightsCost sensitive RAG and chat.Docs · Pricing · Best practices | Cohere | August 2024 | $0.15 | · | $0.60 | 128K | |
| Command R7B Open weightsHigh volume lightweight generation.Docs · Pricing · Best practices | Cohere | December 2024 | $0.0375 | · | $0.15 | 128K | |
| Rerank 3.5Reranking retrieval results in search pipelines.Docs · Pricing · Best practices | Cohere | December 2024 | $2.00 per 1K searches | · | N/A | N/A | |
| Embed 4Enterprise semantic search and multimodal retrieval.Docs · Pricing | Cohere | April 2025 | $0.12 | · | N/A | 128K | |
| Kimi K3 Open weightsComplex multi-step agents, deep research, and large-codebase work where the top tier matters.Docs · Pricing · Best practices | Moonshot | July 2026 | $3.00 | $0.30 | $15.00 | 1M | |
| Kimi K2.7-Code Open weightsAgentic coding, refactors, and tool-driven development inside IDEs and CLI agents.Docs · Pricing · Best practices | Moonshot | June 2026 | $0.95 | $0.19 | $4.00 | 256K | |
| Kimi K2.6 Open weightsTeams wanting a capable open-weight agentic model with generous context at a moderate price.Docs · Pricing · Best practices | Moonshot | April 2026 | $0.95 | $0.16 | $4.00 | 256K | |
| GLM 5.3 PrimeDocs · Pricing | Z.ai | September 2026 | $2.80 | $0.56 | $8.80 | 1M | |
| GLM 5.3 FlashXLatency-sensitive apps.Docs · Pricing · Best practices | Z.ai | September 2026 | $0.37 | $0.09 | $1.25 | 1M | |
| GLM 5.3 Flash Open weightsLow-cost multimodal and agent work.Docs · Pricing · Best practices | Z.ai | August 2026 | $0.15 | $0.03 | $0.50 | 1.25M | |
| GLM 5.3 Open weightsCoding agents and long tasks.Docs · Pricing · Best practices | Z.ai | August 2026 | $1.40 | $0.26 | $4.40 | 1.25M | |
| GLM-5.2 Open weightsFrontier agentic coding and long-context reasoning at a fraction of closed-model cost.Docs · Pricing · Best practices | Z.ai | June 2026 | $1.40 | $0.26 | $4.40 | 1M | |
| GLM-5-TurboAgent workflows that need speed and strong tool calling without the full flagship price.Docs · Pricing · Best practices | Z.ai | 2026 | $1.20 | $0.24 | $4.00 | 200K | |
| GLM-5V-TurboVision-driven coding and UI tasks in agent tools like OpenClaw and Claude Code.Docs · Pricing · Best practices | Z.ai | April 2026 | $1.20 | $0.24 | $4.00 | 200K | |
| GLM-5.1 Open weightsReasoning and coding work on a recent GLM-5 checkpoint.Docs · Pricing · Best practices | Z.ai | April 2026 | $1.40 | $0.26 | $4.40 | 1M | |
| GLM-5 Open weightsSelf-hosting and research on Zhipu's frontier open base.Docs · Pricing · Best practices | Z.ai | February 2026 | Open | · | Open | 1M | |
| GLM-4.7 Open weightsCost-conscious coding and agent tasks that still want deep reasoning.Docs · Pricing · Best practices | Z.ai | December 2025 | $0.60 | $0.11 | $2.20 | 200K | |
| GLM-4.7-FlashHigh-volume, latency-sensitive tasks and prototyping at no token cost.Docs · Pricing · Best practices | Z.ai | December 2025 | Free | · | Free | 128K | |
| GLM-4.6 Open weightsSolid general coding and reasoning where you want a proven, cheaper checkpoint.Docs · Pricing · Best practices | Z.ai | September 2025 | $0.60 | $0.11 | $2.20 | 200K | |
| GLM-4.6V Open weightsMultimodal reasoning and document or UI understanding with tool use.Docs · Pricing · Best practices | Z.ai | December 2025 | $0.30 | $0.05 | $0.90 | 128K | |
| GLM-4.6V-Flash Open weightsOn-device or high-volume multimodal tasks at no token cost.Docs · Pricing · Best practices | Z.ai | December 2025 | Free | · | Free | 128K | |
| GLM-4.5 Open weightsGeneral coding and reasoning on a stable, widely supported checkpoint.Docs · Pricing · Best practices | Z.ai | July 2025 | $0.60 | $0.11 | $2.20 | 131K | |
| GLM-4.5-Air Open weightsBudget-sensitive apps that still want GLM-4.5-class output.Docs · Pricing · Best practices | Z.ai | July 2025 | $0.20 | $0.03 | $1.10 | 131K | |
| GLM-4.5-XCases that need GLM-4.5's strongest, fastest responses and will pay for it.Docs · Pricing · Best practices | Z.ai | July 2025 | $2.20 | · | $8.90 | 131K | |
| GLM-4.5-FlashPrototyping and high-volume light tasks at no token cost.Docs · Pricing · Best practices | Z.ai | July 2025 | Free | · | Free | 131K | |
| GLM-4.5V Open weightsMultimodal understanding on a proven, lower-cost checkpoint.Docs · Pricing · Best practices | Z.ai | August 2025 | $0.60 | $0.11 | $1.80 | 131K | |
| MiniMax-M3 Open weightsFrontier coding, agentic execution, and multimodal reasoning at low cost.Docs · Pricing · Best practices | MiniMax | June 2026 | $0.30 | $0.06 | $1.20 | 1M | |
| MiniMax-M2.5High-volume coding and agent work where value per token is the priority.Docs · Pricing · Best practices | MiniMax | February 2026 | $0.30 | $0.03 | $1.20 | 204K | |
| MiniMax-M2.1 Open weightsSelf-hosting or API use where you want an updated open M2-class model.Docs · Pricing · Best practices | MiniMax | December 2025 | $0.30 | $0.03 | $1.20 | 204K | |
| MiniMax-M2 Open weightsCost-efficient agentic coding at scale on an Apache 2.0 open model.Docs · Pricing · Best practices | MiniMax | October 2025 | $0.26 | · | $1.02 | 204K | |
| MiniMax-M1 Open weightsLong-context reasoning and software tasks that need a big open thinking model.Docs · Pricing · Best practices | MiniMax | June 2025 | $0.40 | · | $2.20 | 1M | |
| MiniMax-Text-01 Open weightsExtreme long-context research and self-hosting on an open linear-attention model.Docs · Pricing | MiniMax | January 2025 | Open | · | Open | 4M | |
| MiniMax-VL-01 Open weightsOpen multimodal research and self-hosting alongside MiniMax-Text-01.Docs · Pricing | MiniMax | January 2025 | Open | · | Open | 1M | |
| Hailuo 02Text-to-video and image-to-video creation for creators and products.Docs · Pricing | MiniMax | 2025 | See pricing | · | See pricing | N/A | |
| Speech 2.5Voiceover, agents, and apps needing natural multilingual speech.Docs · Pricing | MiniMax | 2026 | See pricing | · | See pricing | N/A | |
| Music 2.6Music creation for content, games, and prototyping.Docs · Pricing · Best practices | MiniMax | 2026 | See pricing | · | See pricing | N/A | |
| Hy4 preview Open weightsTrying Tencent's next flagship before general release.Docs · Pricing · Best practices | Tencent | August 2026 | $0.75 | $0.0378 | $2.25 | 1M | |
| Hunyuan Hy3 Open weightsCost-efficient frontier reasoning and agent work, positioned against other top open models.Docs · Pricing · Best practices | Tencent | July 2026 | $0.0825 | $0.0206 | $0.33 | 256K | |
| Hunyuan-TurboSHigh-throughput general chat and assistant tasks that need speed and price efficiency.Docs · Pricing | Tencent | 2025 | $0.11 | · | $0.28 | 256K | |
| Hunyuan-T1Deliberate reasoning tasks in math, code, and analysis.Docs · Pricing | Tencent | March 2025 | See pricing | · | See pricing | 64K | |
| Hunyuan-Large Open weightsSelf-hosting and research on a large open Tencent base.Docs · Pricing | Tencent | November 2024 | Open | · | Open | 256K | |
| Hunyuan-A13B Open weightsLocal deployment and cost-sensitive inference on an open model.Docs · Pricing | Tencent | 2025 | Open | · | Open | 256K | |
| Hunyuan-StandardGeneral assistant and content work that needs long context at a moderate price.Docs · Pricing | Tencent | 2024 | See pricing | · | See pricing | 256K | |
| Hunyuan-LitePrototyping and high-volume light workloads at no token cost.Docs · Pricing | Tencent | 2024 | Free | · | Free | 256K | |
| Hunyuan-VisionMultimodal understanding of documents, charts, and photos.Docs · Pricing | Tencent | 2024 | See pricing | · | See pricing | 32K | |
| HunyuanImage 3.0 Open weightsOpen text-to-image generation and research.Docs · Pricing · Best practices | Tencent | 2026 | Open | · | Open | N/A | |
| HunyuanVideo Open weightsOpen text-to-video and image-to-video generation and research.Docs · Pricing | Tencent | December 2024 | Open | · | Open | N/A | |
| Hunyuan3D 2.5 Open weightsGame, product, and design workflows that need open 3D asset generation.Docs · Pricing | Tencent | 2025 | Open | · | Open | N/A |
Cached input is the price of prompt tokens read from the provider's cache, often a tenth of the full price. Best practices links go to the lab's own guide for that model or its family. Prices are checked every day from each lab's own pricing page (or the lab's own API endpoint), and within about half an hour of news of a price change or a new model; a change shows here with the old price for 90 days. Found a wrong price? Tell us.