Which AI model is cheapest right now?
Prices per 1M tokens, context windows and input types for 7,508 model offers from 199 providers, read from a public open catalogue every 6 hours.
51–100 of 6,826
- Gemma 4 E2B ITAmazon BedrockInput / 1M$0.04Output / 1M$0.08Context131.1K
- L3 8B Stheno V3.2NovitaAIInput / 1M$0.05Output / 1M$0.05Context8.2K
- Qwen/Qwen2.5-7B-InstructSiliconFlowInput / 1M$0.05Output / 1M$0.05Context33K
- Qwen/Qwen2.5-7B-InstructSiliconFlow (China)Input / 1M$0.05Output / 1M$0.05Context33K
- Sao10k L3 8B LunarisNovitaAIInput / 1M$0.05Output / 1M$0.05Context8.2K
- Qwen-VL OCRAlibaba (China)Input / 1M$0.043Output / 1M$0.072Context34.1K
- Qwen3.7 FlashAlibaba (China)Input / 1M$0.03Output / 1M$0.119Context1M
- LFM2 24B A2BPioneerInput / 1M$0.03Output / 1M$0.12Context32.8K
- LFM2-24B-A2BTogether AIInput / 1M$0.03Output / 1M$0.12Context32.8K
- Solar Pro 4LLMTRInput / 1M$0.03Output / 1M$0.12Context524.3K
- Solar Pro 4NanoGPTInput / 1M$0.03Output / 1M$0.12Context524.3K
- Solar Pro 4 ThinkingNanoGPTInput / 1M$0.03Output / 1M$0.12Context524.3K
- Mistral Nemo 12B InstructInferenceInput / 1M$0.038Output / 1M$0.10Context16K
- Qwen TurboAlibaba (China)Input / 1M$0.044Output / 1M$0.087Context1M
- Qwen TurboOfoxInput / 1M$0.043Output / 1M$0.09Context128K
- DeepSeek R1 Distill Llama 70BHeliconeInput / 1M$0.03Output / 1M$0.13Context128K
- gpt-oss-20bCoreWeaveInput / 1M$0.03Output / 1M$0.13Context131.1K
- Llama 3.2 11B Vision InstructInferenceInput / 1M$0.055Output / 1M$0.055Context16K
- Manta Flash 1.0NanoGPTInput / 1M$0.02Output / 1M$0.16Context16.4K
- Manta Mini 1.0NanoGPTInput / 1M$0.02Output / 1M$0.16Context8.2K
- Qwen 3.7 FlashVercel AI GatewayInput / 1M$0.03Output / 1M$0.13Context991K
- Qwen3.7 FlashAlibabaInput / 1M$0.03Output / 1M$0.13Context1M
- Qwen3.7 FlashEmpirioLabs AIInput / 1M$0.03Output / 1M$0.13Context1M
- Qwen3.7 FlashKilo GatewayInput / 1M$0.03Output / 1M$0.13Context1M
- Qwen3.7 FlashDevPass (LLM Gateway)Input / 1M$0.03Output / 1M$0.13Context983.6K
- Qwen3.7 FlashNanoGPTInput / 1M$0.03Output / 1M$0.13Context991.8K
- Qwen3.7 FlashOrcaRouterInput / 1M$0.03Output / 1M$0.13Context1M
- Qwen3.7 Flash (Alibaba Cloud)LLM GatewayInput / 1M$0.03Output / 1M$0.13Context983.6K
- Qwen3.7 Flash ThinkingNanoGPTInput / 1M$0.03Output / 1M$0.13Context983.6K
- DeepSeek V4 Flash 0731engyInput / 1M$0.045Output / 1M$0.09Context1M
- Meta Llama 3.1 8B InstantHeliconeInput / 1M$0.05Output / 1M$0.08Context131.1K
- DeepSeek R1 Distill Llama 70BFastRouterInput / 1M$0.03Output / 1M$0.14Context131.1K
- GPT OSS 20BDeep InfraInput / 1M$0.03Output / 1M$0.14Context131.1K
- GPT OSS 20BVercel AI GatewayInput / 1M$0.03Output / 1M$0.14Context131.1K
- GPT OSS 20B (Deep Infra)Eden AIInput / 1M$0.03Output / 1M$0.14Context131.1K
- GPT-OSS 20BIO.NETInput / 1M$0.03Output / 1M$0.14Context64K
- Llama 3.1 8BGroqInput / 1M$0.05Output / 1M$0.08Context131.1K
- Mistral: Mistral Small 3Kilo GatewayInput / 1M$0.05Output / 1M$0.08Context32.8K
- GPT OSS 120BDevPass (LLM Gateway)Input / 1M$0.032Output / 1M$0.14Context131.1K
- GPT OSS 120B (Runware)LLM GatewayInput / 1M$0.032Output / 1M$0.14Context131.1K
- Inference.net: Schematron V2 TurboKilo GatewayInput / 1M$0.03Output / 1M$0.15Context128K
- Llama-3.1-8B-InstructHugging FaceInput / 1M$0.06Output / 1M$0.06Context131.1K
- Mistral Devstral Small 2505NanoGPTInput / 1M$0.06Output / 1M$0.06Context32.8K
- Qwen/Qwen3-8BSiliconFlowInput / 1M$0.06Output / 1M$0.06Context131K
- Qwen/Qwen3-8BSiliconFlow (China)Input / 1M$0.06Output / 1M$0.06Context131K
- Schematron V2 TurboNanoGPTInput / 1M$0.03Output / 1M$0.15Context128K
- Schematron V2 TurboVercel AI GatewayInput / 1M$0.03Output / 1M$0.15Context128K
- Qwen TurboOfoxInput / 1M$0.05Output / 1M$0.09Context128K
- AutoGLM-Phone-9B-MultilingualNovitaAIInput / 1M$0.035Output / 1M$0.138Context65.5K
- Qwen3 8BNovitaAIInput / 1M$0.035Output / 1M$0.138Context128K
Price changes
Models whose price changed in the catalogue, newest first. We compare prices every day.
- DeepSeek: DeepSeek V4 Flash LatestKilo GatewayInput / 1M$0.013$0.005775▼ 55%Output / 1M$1.60$1.47▼ 8%Changed: Oct 2, 2026, 7:19 PM UTC
- DeepSeek: DeepSeek Flash LatestKilo GatewayInput / 1M$0.015$0.02▲ 33%Output / 1M$1.20$0.60▼ 50%Changed: Oct 2, 2026, 7:19 PM UTC
- Claude Sonnet 5.5Venice AIInput / 1M$3.75$2.50▼ 33%Output / 1M$18.75$12.50▼ 33%Changed: Oct 2, 2026, 5:25 PM UTC
- Claude Sonnet 5Venice AIInput / 1M$3.00$2.50▼ 17%Output / 1M$15.00$12.50▼ 17%Changed: Oct 2, 2026, 5:25 PM UTC
- GLM 4.7 ThinkingNanoGPTInput / 1M$0.20$0.40▲ 100%Output / 1M$0.80$1.93▲ 141%Changed: Oct 2, 2026, 5:25 PM UTC
- GLM 4.7NanoGPTInput / 1M$0.20$0.40▲ 100%Output / 1M$0.80$1.93▲ 141%Changed: Oct 2, 2026, 5:25 PM UTC
- GPT-OSS 20BMerge GatewayInput / 1M$0.04Output / 1M$0.20$0.15▼ 25%Changed: Oct 2, 2026, 5:25 PM UTC
- GPT-OSS 120BMerge GatewayInput / 1M$0.09$0.05▼ 44%Output / 1M$0.36$0.25▼ 31%Changed: Oct 2, 2026, 5:25 PM UTC
- Z.ai: GLM Flash LatestKilo GatewayInput / 1M$0.026Output / 1M$0.625$0.90▲ 44%Changed: Oct 2, 2026, 5:25 PM UTC
- Hy4 previewKilo GatewayInput / 1M$0.834$0.751▼ 10%Output / 1M$2.50$2.25▼ 10%Changed: Oct 2, 2026, 5:25 PM UTC
- Hy3Kilo GatewayInput / 1M$0.13$0.083▼ 37%Output / 1M$0.53$0.33▼ 38%Changed: Oct 2, 2026, 5:25 PM UTC
- MoonshotAI: Kimi LatestKilo GatewayInput / 1M$1.39$1.34▼ 4%Output / 1M$13.00Changed: Oct 2, 2026, 5:25 PM UTC
- Nano BananaKilo GatewayInput / 1M$0.15$0.30▲ 100%Output / 1M$1.25$2.50▲ 100%Changed: Oct 2, 2026, 5:25 PM UTC
- DeepSeek: DeepSeek V4 Flash LatestKilo GatewayInput / 1M$0.003825$0.013▲ 235%Output / 1M$1.60Changed: Oct 2, 2026, 5:25 PM UTC
- DeepSeek: DeepSeek Flash LatestKilo GatewayInput / 1M$0.03$0.015▼ 50%Output / 1M$0.75$1.20▲ 60%Changed: Oct 2, 2026, 5:25 PM UTC
- Llama-3.3-70B-Instruct (Scaleway)Eden AIInput / 1M$1.02$1.01▼ 1%Output / 1M$1.02$1.01▼ 1%Changed: Oct 2, 2026, 5:25 PM UTC
- GPT OSS 120B (Scaleway)Eden AIInput / 1M$0.169$0.168▼ 1%Output / 1M$0.678$0.674▼ 1%Changed: Oct 2, 2026, 5:25 PM UTC
- DeepSeek V4 Flash 0731 (Scaleway)Eden AIInput / 1M$0.452$0.449▼ 1%Output / 1M$0.904$0.898▼ 1%Changed: Oct 2, 2026, 5:25 PM UTC
- GPT OSS 120B (IONOS)Eden AIInput / 1M$0.169$0.168▼ 1%Output / 1M$0.734$0.73▼ 1%Changed: Oct 2, 2026, 5:25 PM UTC
- Llama-3.3-70B-Instruct (IONOS)Eden AIInput / 1M$0.734$0.73▼ 1%Output / 1M$0.734$0.73▼ 1%Changed: Oct 2, 2026, 5:25 PM UTC
- Ministral 3 14B (Infomaniak)Eden AIInput / 1M$0.339$0.337▼ 1%Output / 1M$0.452$0.449▼ 1%Changed: Oct 2, 2026, 5:25 PM UTC
- DeepSeek V4 ProDeepSeekInput / 1M$0.435$0.66▲ 52%Output / 1M$0.87$1.98▲ 128%Changed: Oct 2, 2026, 5:25 PM UTC
- MiMo V2.6 FlashVercel AI GatewayInput / 1M$0.14$0.04▼ 71%Output / 1M$0.28$1.28▲ 357%Changed: Oct 2, 2026, 6:26 AM UTC
- GPT-5.5 (EU)RequestyInput / 1M$5.00$5.50▲ 10%Output / 1M$30.00$33.00▲ 10%Changed: Oct 2, 2026, 6:26 AM UTC
- GPT-5.4 (EU)RequestyInput / 1M$2.50$2.75▲ 10%Output / 1M$15.00$16.50▲ 10%Changed: Oct 2, 2026, 6:26 AM UTC
- DeepSeek V4.1 Flash (Consensus Protocol)LLM GatewayInput / 1M$0.15Output / 1M$0.60$0.55▼ 8%Changed: Oct 2, 2026, 6:26 AM UTC
- DeepSeek V4.1 FlashDevPass (LLM Gateway)Input / 1M$0.15Output / 1M$0.60$0.55▼ 8%Changed: Oct 2, 2026, 6:26 AM UTC
- Z.ai: GLM LatestKilo GatewayInput / 1M$0.133$0.12▼ 10%Output / 1M$4.00Changed: Oct 2, 2026, 6:26 AM UTC
- Z.ai: GLM Flash LatestKilo GatewayInput / 1M$0.02$0.026▲ 31%Output / 1M$0.248$0.625▲ 153%Changed: Oct 2, 2026, 6:26 AM UTC
- MoonshotAI: Kimi LatestKilo GatewayInput / 1M$0.58$1.39▲ 140%Output / 1M$10.00$13.00▲ 30%Changed: Oct 2, 2026, 6:26 AM UTC
- Kimi K3Kilo GatewayInput / 1M$0.58$2.70▲ 366%Output / 1M$10.00$13.50▲ 35%Changed: Oct 2, 2026, 6:26 AM UTC
- Kimi K2.6Kilo GatewayInput / 1M$0.65$0.434▼ 33%Output / 1M$3.41$1.83▼ 46%Changed: Oct 2, 2026, 6:26 AM UTC
- DeepSeek: DeepSeek V4 Flash LatestKilo GatewayInput / 1M$0.004455$0.003825▼ 14%Output / 1M$0.131$1.60▲ 1124%Changed: Oct 2, 2026, 6:26 AM UTC
- DeepSeek: DeepSeek Pro LatestKilo GatewayInput / 1M$0.291$0.132▼ 55%Output / 1M$3.50$0.396▼ 89%Changed: Oct 2, 2026, 6:26 AM UTC
- DeepSeek: DeepSeek Flash LatestKilo GatewayInput / 1M$0.016$0.03▲ 93%Output / 1M$0.396$0.75▲ 89%Changed: Oct 2, 2026, 6:26 AM UTC
- DeepSeek V4.1 FlashCharm HyperInput / 1M$0.30$0.33▲ 10%Output / 1M$1.20$1.31▲ 9%Changed: Oct 2, 2026, 6:26 AM UTC
- DeepSeek Flash LatestFireworks AIInput / 1M$0.22$0.30▲ 36%Output / 1M$0.66$1.20▲ 82%Changed: Oct 2, 2026, 6:26 AM UTC
- DeepSeek V4.1 FlashFireworks AIInput / 1M$0.22$0.30▲ 36%Output / 1M$0.66$1.20▲ 82%Changed: Oct 2, 2026, 6:26 AM UTC
- Llama-3.3-70B-Instruct (Scaleway)Eden AIInput / 1M$1.02$1.02▼ 1%Output / 1M$1.02$1.02▼ 1%Changed: Oct 2, 2026, 6:26 AM UTC
- GPT OSS 120B (Scaleway)Eden AIInput / 1M$0.17$0.169▼ 1%Output / 1M$0.681$0.678▼ 1%Changed: Oct 2, 2026, 6:26 AM UTC
How this works
- Every price, limit and date comes from models.dev, an open-source (MIT-licensed) catalogue of AI models kept in public on GitHub and compiled from providers’ official pricing and documentation pages.
- We read it at most every 6 hours. Providers change prices without notice, so confirm on the provider’s own page (linked on each model) before you rely on a number.
- Prices are US dollars per 1 million tokens, exactly as the catalogue lists them. The blended price, (3 × input + output) ÷ 4, is our own estimate, used only to sort and compare.
- Anything the catalogue does not state is shown as “—”; we never guess. Flat-fee subscription plans, local-only runtimes and deprecated models are left out, and $0 offers are hidden unless you include them.
- No benchmark scores yet. When we add them, they will come only from official sources and be labelled as such.
Sources: models.dev · api.json · GitHub · MIT License
Wrong price, or a provider that wants its listing corrected or removed? Email ceo@uniholix.com. We reply within 5 business days.