Intelligence des modèles d'IA

Tarifs · 2026-11-17

Tarifs des LLM

Coût par jeton chez les grands fournisseurs, normalisé en USD par 1M de jetons.

Cheapest LLM APIs

AI APIs ranked by input + output token cost.

OpenAI API Pricing

All OpenAI model prices in one table — GPT-5, GPT-5 Mini, embeddings and more.

Anthropic Claude Pricing

All Anthropic Claude prices — Opus, Sonnet, Haiku and prompt caching costs.

À propos de cette liste

Almost every commercial LLM provider charges separately for input tokens (the prompt you send) and output tokens (the response you get back). Output tokens typically cost 3–5× more than input tokens because generation is autoregressive — each token depends on the previous one and cannot be batched as efficiently.

The table below normalises all prices to USD per 1 million tokens. This is the industry-standard unit; per-1K or per-token rates are easy to misread by 1000×. The "Total" column sums input + output for a quick apples-to-apples comparison, but your real cost depends heavily on your input/output ratio.

What the table does NOT show

  • Prompt caching discounts — Anthropic and OpenAI offer cache_read rates 5–10× cheaper than standard input. If you reuse a long system prompt, caching dominates total cost.
  • Tiered pricing above 200K tokens — Google and Anthropic charge a premium for very long inputs. Each model detail page shows the >200K tier when applicable.
  • Reasoning tokens — thinking models (o-series, Claude Extended Thinking, DeepSeek R1) bill internal reasoning at output rates, often 2–3× the visible answer length.
  • Volume discounts and prepaid credits — most providers offer 10–30% off at scale. These are not reflected here.
  • Audio / image surcharges — multimodal inputs have separate per-image or per-second rates.

How to use this page

  1. Find the cheapest model that meets your context window and capability requirements (tool calling, structured output, vision).
  2. Open its detail page to check cache pricing, output limits and provider availability.
  3. Use the pricing calculator to estimate monthly cost for your actual token volume.
  4. Compare 2–3 finalists side-by-side on a comparison page.

Prices are refreshed daily. Models showing "Unknown" do not publish a public per-token rate — this usually means enterprise-only or invite-gated access. We deliberately do not show them as $0 to avoid misleading rankings.

#ModelVendorInput / 1MOutput / 1MTotalContext
1Qwen3.8 27BAlibaba (Qwen)$0.001$0.001$0.0021.02M
2Ambient Smallambient$0.001$0.001$0.002101K
3Multi-QA-mpnet-base-dot-v1digitalocean$0.009Unknown$0.009512
4All-MiniLM-L6-v2digitalocean$0.009Unknown$0.009256
5Qwen3 Embedding 8BAlibaba (Qwen)$0.010Unknown$0.01033K
6Qwen3 Embedding 4BAlibaba (Qwen)$0.010Unknown$0.01033K
7BGE Reranker v2 M3digitalocean$0.010Unknown$0.0108K
8BGE Reranker V2 M3novita-ai$0.010Unknown$0.0108K
9BGE M3novita-ai$0.010Unknown$0.0108K
10Nomic Embed Text v1.5tinfoil$0.016Unknown$0.0168K
11Llama 3.2 1B InstructMeta$0.010$0.010$0.02060K
12Qwen3 Embedding 0.6BAlibaba (Qwen)$0.010$0.010$0.02033K
13Llama Prompt Guard 2 22MMeta$0.010$0.010$0.020512
14Prompt Guard 2 86MMeta$0.010$0.010$0.020512
15text-embedding-3-smallOpenAI$0.020Unknown$0.0208K
16text-embedding-3-smallsap-ai-core$0.020Unknown$0.0208K
17amazon--titan-embed-textsap-ai-core$0.020Unknown$0.0208K
18BGE M3digitalocean$0.020Unknown$0.0208K
19E5 Large v2digitalocean$0.020Unknown$0.020512
20text-embedding-3-smallazure-cognitive-services$0.020Unknown$0.0208K
21text-embedding-3-smallazure$0.020Unknown$0.0208K
22DeepSeek V4 Flash 0731DeepSeek$0.006$0.019$0.0251.05M
23GLiNER 2.5 Multipioneer$0.030Unknown$0.0304K
24GLiNER2 Multipioneer$0.030Unknown$0.0308K
25GLiNER2 Privacy Filter PII (Multi)pioneer$0.030Unknown$0.0308K
26GLiNER 2.5 Basepioneer$0.030Unknown$0.0304K
27GLiNER2 Largepioneer$0.030Unknown$0.0308K
28GLiNER2 Multi Largepioneer$0.030Unknown$0.0308K
29GLiNER-2.5-Decidepioneer$0.030Unknown$0.0308K
30GLiNER2 Basepioneer$0.030Unknown$0.0302K
31Llama 3.2 3B InstructMeta$0.020$0.020$0.040131K
32Summarizeopenrouter$0.040Unknown$0.0401M
33Extractopenrouter$0.040Unknown$0.0401M
34PaddleOCR VLnovita-ai$0.020$0.020$0.04016K
35Llama-3.1-8B-InstructMeta$0.025$0.025$0.050131K
36Summarize-1vercel$0.050Unknown$0.0501M
37Extract-1vercel$0.050Unknown$0.0501M
38Meta Llama 3.1 8B Instruct TurboMeta$0.020$0.030$0.050128K
39ScaleDown: Summarize-1kilo$0.050Unknown$0.0501M
40ScaleDown: Summarizekilo$0.050Unknown$0.0501M
41ScaleDown: Extract-1kilo$0.050Unknown$0.0501M
42ScaleDown: Extractkilo$0.050Unknown$0.0501M
43Mistral NemoMistral$0.029$0.030$0.059131K
44Mistral NemoMistral$0.020$0.040$0.06016K
45DeepSeek OCR 2DeepSeek$0.030$0.030$0.0604K
46Ling 3.0 Tinynano-gpt$0.015$0.045$0.060262K
47Ling 3.0 Tiny Thinkingnano-gpt$0.015$0.045$0.060262K
48Ling 3.0 Tinyopenrouter$0.015$0.045$0.060262K
49inclusionAI: Ling 3.0 Tinykilo$0.015$0.045$0.060262K
50Ministral 3Bazure-cognitive-services$0.040$0.040$0.080128K
51Ministral 3Bazure$0.040$0.040$0.080128K
52Ling 3.0 Flashvercel$0.021$0.063$0.084256K
53Ling 3.0 Flashopenrouter$0.021$0.063$0.084262K
54gpt-oss-20bOpenAI$0.015$0.075$0.090131K
55Gemma 3 4B ITGoogle$0.040$0.050$0.090131K
56DeepSeek V4.1 Flash FastDeepSeek$0.010$0.080$0.0901.05M
57Llama 3 8B LunarisMeta$0.040$0.050$0.0908K
58GTE Large (v1.5)digitalocean$0.090Unknown$0.0908K
59LightOnOCR 2 1Bstepping-stone$0.024$0.071$0.09516K
60MinerU2.5 2509 1.2Bstepping-stone$0.024$0.071$0.09516K
61Qwen2.5 7B Instruct (Deep Infra)Alibaba (Qwen)$0.050$0.050$0.10033K
62Mistral EmbedMistral$0.100Unknown$0.1008K
63text-embedding-ada-002OpenAI$0.100Unknown$0.1008K
64text-embedding-ada-002azure-cognitive-services$0.100Unknown$0.1008K
65text-embedding-ada-002azure$0.100Unknown$0.1008K
66GPT OSS 20B (FlexAI, EU)edenai$0.018$0.090$0.108131K
67Llama-3.2-11B-Vision-Instruct (Deep Infra)Meta$0.055$0.055$0.110131K
68Llama-Guard-3-8B (Cloudflare)Meta$0.055$0.055$0.110131K
69Qwen3.5 4BAlibaba (Qwen)$0.040$0.070$0.110262K
70GLM Flash LatestZ.AI / Zhipu$0.040$0.078$0.1181.05M
71BGE Reranker V2 M3stepping-stone$0.119Unknown$0.1198K
72BGE M3stepping-stone$0.119Unknown$0.1198K
73Gemma 4 26B A4B ITGoogle$0.030$0.090$0.120262K
74Nex N2.5 Mininano-gpt$0.025$0.100$0.125262K
75Nex-N2.5-Miniopenrouter$0.025$0.100$0.125262K
76Nex AGI: Nex-N2.5-Minikilo$0.025$0.100$0.125262K
77Granite 4.0 H Microcloudflare-workers-ai$0.017$0.112$0.129131K
78Mistral Small 3Mistral$0.050$0.080$0.13033K
79text-embedding-3-largeOpenAI$0.130Unknown$0.1308K
80Llama 3.1 8BMeta$0.050$0.080$0.130131K
81text-embedding-3-largesap-ai-core$0.130Unknown$0.1308K
82text-embedding-3-largeazure-cognitive-services$0.130Unknown$0.1308K
83Synth 2.5 Flash Previewnano-gpt$0.030$0.100$0.13030K
84text-embedding-3-largeazure$0.130Unknown$0.1308K
85Model Routerazure-cognitive-services$0.140Unknown$0.140200K
86Baichuan M2 32Bnovita-ai$0.070$0.070$0.140131K
87Model Routerazure$0.140Unknown$0.140200K
88Apertus v1.5 8Bstepping-stone$0.024$0.119$0.143131K
89Nemotron 3.5 Lightning 30B A3BNVIDIA$0.025$0.120$0.1451.05M
90DeepSeek-V4.1-FlashDeepSeek$0.030$0.120$0.1501M
91DeepSeek V4.1 Flash 1017DeepSeek$0.030$0.120$0.1501.05M
92Gemini Embedding 001Google$0.150Unknown$0.1502K
93LFM2-24B-A2Btogetherai$0.030$0.120$0.15033K
94Gemma 3 12B ITGoogle$0.050$0.100$0.150131K
95Qwen3.7 FlashAlibaba (Qwen)$0.030$0.130$0.1601M
96DeepSeek R1 Distill Llama 70BMeta$0.030$0.130$0.16033K
97Phi-4Microsoft$0.056$0.112$0.168128K
98amazon--nova-microsap-ai-core$0.035$0.140$0.175128K
99Nova Microvercel$0.035$0.140$0.175128K
100Nova Micro (US)edenai$0.035$0.140$0.175128K

Showing top 100 of 1688. Use the full directory to see the rest.

Frequently asked questions

Why are output tokens more expensive than input tokens?

Generation is autoregressive — each output token requires a full forward pass conditioned on all previous tokens, which cannot be batched as efficiently as reading input tokens in parallel. The 3–5× premium most providers charge reflects this compute asymmetry.

What does 'per 1M tokens' mean in practice?

One million tokens is roughly 750,000 English words or 3,000 pages of standard text. A typical chatbot request uses 1,000–5,000 input tokens and 200–1,000 output tokens, so 1M tokens represents hundreds to thousands of requests depending on your workload.

Why are some models showing 'Unknown' instead of a price?

We deliberately do not coerce missing data to $0. 'Unknown' means the provider does not publish a public per-token rate — often models behind enterprise sales or invite-only access. Treating Unknown as free would push paid-but-unpriced models to the top of every cheap list.

How often do these prices change?

Vendor list-price moves are typically picked up within hours of an announcement, and our pipeline re-syncs daily. Each change is written to /changelog so you can audit historical pricing over time.

Does this include prompt caching or batch discounts?

No. The table shows standard headline rates only. Prompt caching (Anthropic, OpenAI) can reduce input cost by 50–90%. Batch API discounts (OpenAI) offer ~50% off for non-real-time workloads. Both are shown on each model's detail page.

Last updated:

Prices in USD per 1M tokens. Unknown means the provider does not publish per-token pricing.

Pricing and capabilities are refreshed daily and reconciled against each provider's official documentation. Always verify critical production decisions with the provider directly.