Mistral API Pricing (November 2026)
Mistral API prices run from $0.10 to $7.50 per million tokens across 8 current models. The cheapest is Ministral 3 3B at $0.10 input and $0.10 output per million tokens. The flagship, Mistral Large 3, costs $0.50 input and $1.50 output per million tokens. Every rate below was read on Mistral AI's own pricing page, on the date shown in its row.
| Model | Input /M | Cache read /M | Output /M | Batch in /M | Batch out /M | Context | Status | Checked |
|---|---|---|---|---|---|---|---|---|
| Mistral Small 4 | $0.15 | $0.015−90% | $0.60 | $0.075 | $0.30 | 256K | Current | 2026-10-09 |
| Ministral 3 14B | $0.20 | $0.02−90% | $0.20 | $0.10 | $0.10 | 128K | Current | 2026-10-09 |
| Ministral 3 8B | $0.15 | $0.015−90% | $0.15 | $0.075 | $0.075 | 128K | Current | 2026-10-09 |
| Mistral Medium 3.5 | $1.50 | $0.15−90% | $7.50 | $0.75 | $3.75 | documented elsewhere — not on pricing page | Current | 2026-10-09 |
| Ministral 3 3B | $0.10 | $0.01−90% | $0.10 | $0.05 | $0.05 | 128K | Current | 2026-10-09 |
| Mistral Large 3 | $0.50 | $0.05−90% | $1.50 | $0.25 | $0.75 | documented elsewhere — not on pricing page | Flagship | 2026-10-09 |
| Codestral | $0.30 | $0.03−90% | $0.90 | — | — | documented elsewhere — not on pricing page | Current | 2026-10-09 |
Mistral Large 4 mistral-large-4 | $1.36 | $0.14−90% | $4.18 | $0.68 | $2.09 | 1M | Preview | 2026-10-20 |
US dollars per million tokens, list rates on Mistral AI's own API. Cache read is the price of a cached input token. Batch is the asynchronous API, where the vendor offers one. The date in each row is when we last read it on the vendor's pricing page; it links to that page.
Mistral API monthly cost estimates
What light use (10M input + 2M output tokens a month), medium use (100M input + 20M output tokens a month), heavy use (1B input + 200M output tokens a month) costs on each current model, at list price with no caching and no batch discount.
| Model | Input / output per 1M | Light · 10M + 2M | Medium · 100M + 20M | Heavy · 1B + 200M |
|---|---|---|---|---|
| Ministral 3 3B | $0.10 / $0.10 | $1.20 | $12.00 | $120 |
| Ministral 3 8B | $0.15 / $0.15 | $1.80 | $18.00 | $180 |
| Mistral Small 4 | $0.15 / $0.60 | $2.70 | $27.00 | $270 |
| Ministral 3 14B | $0.20 / $0.20 | $2.40 | $24.00 | $240 |
| Codestral | $0.30 / $0.90 | $4.80 | $48.00 | $480 |
| Mistral Large 3 | $0.50 / $1.50 | $8.00 | $80.00 | $800 |
| Mistral Medium 3.5 | $1.50 / $7.50 | $30.00 | $300 | $3,000 |
Prompt caching and the Batch API cut these figures; the next sections show by how much. For your own token counts, use the LLM API cost calculator.
Mistral API price history
Mistral API price changes and releases
Major releases, funding, and EU-infrastructure moves. Sourced from official Mistral announcements + press coverage, verified at publication.
$1.36/$4.18 with a two-week 50% launch sale; weights promised by the end of October$0.15/$0.60 from $0.10/$0.30, unannouncedRetired Mistral models and their replacements
24 Mistral models no longer answer on Mistral AI's API. The price is the last one Mistral AI listed; the replacement is the current end of the vendor's own succession chain.
| Model | Status | Last price in / out per 1M | Use instead |
|---|---|---|---|
| Codestral 2501 | Retired Nov 30, 2025 | $0.30 / $0.90 | Codestral |
| Codestral 2405 | Retired Jun 16, 2025 | $0.20 / $0.60 | Codestral |
| Pixtral 12B | Retired Dec 31, 2025 | $0.15 / $0.15 | Ministral 3 14B |
| Mixtral 8x22B | Retired Mar 30, 2025 | $2.00 / $6.00 | Mistral Small 4 |
| Mixtral 8x7B | Retired Mar 30, 2025 | $0.70 / $0.70 | Mistral Small 4 |
| Ministral 8B | Retired Dec 31, 2025 | $0.10 / $0.10 | Ministral 3 8B |
| Mistral 7B | Retired Mar 30, 2025 | $0.25 / $0.25 | Ministral 3 8B |
| Mistral Small 3.2 | Retired Jul 31, 2026 | $0.10 / $0.30 | Mistral Small 4 |
| Mistral Medium 3.1 | Retired Aug 31, 2026 | $1.50 / $7.50 | Mistral Medium 3.5 |
| Ministral 3B | Retired Dec 31, 2025 | $0.04 / $0.04 | Ministral 3 3B |
| Mistral Large 2.1 | Retired May 31, 2026 | $2.00 / $6.00 | Mistral Medium 3.5 |
| Mistral Large 2 | Retired Mar 30, 2025 | $2.00 / $6.00 | Mistral Large 3 |
| Devstral 2 | Retired Jul 31, 2026 | $0.40 / $2.00 | Mistral Medium 3.5 |
| Devstral Small 2 | Retired Mar 31, 2026 | $0.10 / $0.30 | Mistral Small 4 |
| Leanstral 1.5 | Retired Sep 30, 2026 | Free / Free | — |
| Mistral Large 1 | Retired Jun 16, 2025 | $4.00 / $12.00 | Mistral Large 3 |
| Mistral Medium 1 | Retired Jun 16, 2025 | $2.70 / $8.10 | Mistral Medium 3.5 |
| Mistral Small 1 | Retired Jun 16, 2025 | $1.00 / $3.00 | Mistral Small 4 |
| Pixtral Large | Retired May 31, 2026 | $2.00 / $6.00 | Mistral Medium 3.5 |
| Magistral Medium | Retired Jul 31, 2026 | $2.00 / $5.00 | Mistral Medium 3.5 |
| Magistral Small | Retired Jul 31, 2026 | $0.50 / $1.50 | Mistral Small 4 |
| Codestral Mamba 7B | Retired Jun 6, 2025 | $0.25 / $0.25 | Codestral |
| Mistral NeMo 12B | Retired Jul 31, 2026 | $0.15 / $0.15 | Ministral 3 8B |
| Leanstral | Retired Jun 30, 2026 | Free / Free | — |
console.mistral.ai — direct API access, dev console, fine-tuning, and billing. Full model catalog including premier and free rows. Batch API offers −50% on async jobs.
CLOUD · MULTIMistral models are resold via AWS Bedrock, Azure AI Foundry, and Google Vertex AI Model Garden — useful for teams that bill through an existing cloud contract.
OPEN-WEIGHT · SELF-HOSTOpen-weight tiers (Small, Ministral, Devstral Small, Mixtral) ship under Apache 2.0 on Hugging Face — fully self-hostable, no per-token fee. Premier models stay API-only.
CONSUMER · WEBchat.mistral.ai — Free, Pro, and Enterprise tiers. Web, iOS, Android. Le Chat Enterprise adds EU data residency, agents, and memory for regulated teams.
SOVEREIGN · INFRAEU-sovereign GPU infrastructure built with Nvidia (Blackwell / GB300) — lets European companies own their full AI stack without US or Chinese cloud dependence.
OpenAI
Wider lineup + ecosystem (GPT-5 family, ChatGPT, Azure). Stronger frontier benchmarks. But closed weights, US data residency, and pricier output.
FRONTIER LEADERAnthropic
Leads coding + reasoning benchmarks with Claude. Zero-retention defaults. But no open weights, higher prices, and no EU-native data residency by default.
ENTERPRISE RAGCohere
Enterprise RAG + embeddings focus (Command, Embed, Rerank) with strong retrieval tooling. Smaller consumer presence and a narrower frontier-model story.
Aleph Alpha
German-made, used by EU governments for full data sovereignty. Compliance-strong but smaller scale and weaker public benchmarks than Mistral.
Meta
Llama open weights at huge scale, self-hostable and free to run. But no managed EU-resident API and a US-centric data + governance posture.
Alibaba (Qwen)
The Qwen (Tongyi Qianwen) model family from Alibaba Cloud's Tongyi Lab. First released in 2023, Qwen is the most-downloaded open-weight model family in the world — most tiers ship under Apache 2.0 on Hugging Face and ModelScope, while the proprietary Max tier is served through Alibaba Cloud's Model Studio.
Mistral API pricing FAQ
Practical questions about Mistral AI, open weights, and EU data residency.
Q · 01 Which Mistral model should I start with?
+
$1.5/$7.5) for agentic and coding work, or Mistral Large 3 ($0.5/$1.5) if you want the open-weight flagship. Mistral Large 4 ($1.36/$4.18 list, half that during its launch sale) is the largest model, but it is a public preview and its weights are not out yet. Drop to Mistral Small 4 ($0.15/$0.60) for cheap high-volume multimodal. See the use case picker above. Q · 02 Is Mistral a good choice for EU data residency?
+
Q · 03 Are Mistral models open-weight?
+
Apache 2.0 on Hugging Face — fully self-hostable with no per-token fee. Premier models like Mistral Medium 3.5 and Magistral Medium are API-only. Mistral Large 4 is described as open-weight, but on October 6, 2026 it launched as an API-only public preview; Mistral says the weights follow by the end of October. Check each model card before assuming a license. Q · 04 What's the data retention policy?
+
Q · 05 Is there a free option?
+
$0.1/M (symmetric input/output). Open weights also let you self-host any Apache-2.0 model for compute cost only. Q · 06 Can I get a discount?
+
$0.05/M against $0.50/M. The calculators here apply it at the site-wide 82% cache-hit assumption. Priority-tier (low-latency) APIs run about +75% over list. Self-hosting open weights removes per-token cost entirely. Q · 07 How does Mistral compare to OpenAI, Anthropic, and DeepSeek?
+
Mistral AI was founded in April 2023 in Paris by Arthur Mensch (CEO, ex-Google DeepMind), Guillaume Lample (Chief Scientist) and Timothée Lacroix (CTO) — the latter two from Meta's Llama team. The pitch: build frontier-grade models in Europe with open weights and EU data sovereignty as core principles rather than afterthoughts.
The company is a private French société par actions simplifiée (SAS). Funding has been rapid: a €105M seed in mid-2023, a ~$2B-valuation Series A in late 2023, a ~$6B round in 2024, and a €1.7B Series C in September 2025 led by chip-equipment maker ASML that more than doubled its valuation to roughly $14B. Investors include Andreessen Horowitz, Nvidia, Lightspeed, General Catalyst, DST Global, Index Ventures and Bpifrance.
Mistral's technical signature is its open-weight release strategy: the Mixtral mixture-of-experts models, the Small and Ministral families, and Devstral Small ship under Apache 2.0 licenses on Hugging Face — fully self-hostable. Premier models like Mistral Medium 3.5 and Magistral Medium remain API-only. The lab also builds the Le Chat consumer assistant and the Codestral / Devstral coding line.
Beyond models, Mistral is investing in European AI sovereignty. Mistral Compute, unveiled at VivaTech in June 2025 with Nvidia, provides sovereign GPU infrastructure so EU companies and governments can own their AI stack without depending on US or Chinese cloud providers; the lab raised $830M in debt in 2026 to build a Paris-region data-center cluster.
Versus US labs: Mistral's frontier benchmarks trail OpenAI and Anthropic, and its lineup is narrower. But it is materially cheaper, ships open weights, and is the clearest EU-native, GDPR-first option — a meaningful differentiator for European procurement and regulated industries.