Mistral API pricing
Mistral AI's API models, led by Mistral Large 4 ($0.70 / $2.10), down to the Ministral 3 family at $0.10 to $0.20 per 1M tokens. Batch requests cost 50% less.
Mistral Large 4
The 4th cheapest of 8 current workhorse models (blended 3:1). Cheapest: MiniMax M3, $0.30 / $1.20.
Capability score 155, 19th of 29 current models scored on the Epoch Capabilities Index (range 152–157).
GPT-6 Luna scores level (157) and costs 7× less on a coding agent: $25.98 vs $173 a month.
Source: Epoch Capabilities Index by Epoch AI, CC BY 4.0, read 11 Nov 2026. What the score means
What it costs in practice
Mistral Large 4 at list prices, no batch discount.
| 1,000 chat replies1,500 tokens in and 300 out each | $1.68 |
|---|---|
| Summarising 100 long reports20,000 tokens in and 800 out each | $1.57 |
| A month of a coding agent400 tasks, 85% of input cached | $140 |
The coding-agent example assumes 85% of input is served from cache, typical for agents that resend the same context each step. Work out AI agent cost on Mistral Large 4
Details
mistral-large-4- Labelled a public preview by Mistral.
- Replaces Mistral Large 3 and Mistral Medium 3.5.
Checked 2 Nov 2026 against Mistral pricing and Mistral model card .
Estimate your monthly cost
List prices, no batch discount
| Model | Input / output per 1M | Per month |
|---|---|---|
| Mistral Large 4 | $0.70 / $2.10 | $11.20 |
| Codestral 25.08 | $0.30 / $0.90 | $4.80 |
| Mistral Small 4 | $0.15 / $0.60 | $2.70 |
| Ministral 3 14B | $0.20 / $0.20 | $2.40 |
| Ministral 3 3B | $0.10 / $0.10 | $1.20 |
| Ministral 3 8B | $0.15 / $0.15 | $1.80 |
| MiniMax M3Cheaper alternative · MiniMax | $0.30 / $1.20 | $5.40 |
| Gemini 3.8 FlashSimilar price · Google | $0.75 / $3.75 | $15.00 |
Type amounts like 500K, 25M or 2B.
Price Mistral Large 4 in: Token cost calculator · AI agent cost · Token counter
Cheaper alternative
MiniMax M3 (MiniMax) costs $0.30 / $1.20 per 1M tokens, 50% less than Mistral Large 4 at a 3:1 input-to-output mix. It’s the cheapest current workhorse model from another family. Cheaper doesn’t mean as good: test it on your own tasks.
Compare Mistral Large 4 with MiniMax M3
Similar price from another provider: Gemini 3.8 Flash, $0.75 / $3.75. Compare
Compare Mistral Large 4 with
Side by side: prices, cached and batch rates, and the cost of the same workload.
Other Mistral models
Large 3 and Medium 3.5 were deprecated on 31 October 2026 and retire on 30 April 2027. Mistral retired Nemo, Small 3.2, Devstral 2 and Magistral 1.2 on 31 July 2026.
| Model | Input / 1M | Output / 1M | Cached input | Context | Released |
|---|---|---|---|---|---|
| $0.30 | $0.90 | $0.030 | 256K | Aug 2025 | |
Open weights · Source: Mistral pricing (opens in a new tab) |
$0.15 | $0.60 | $0.015 | 256K | 16 Mar 2026 |
| $0.20 | $0.20 | $0.020 | — | 2 Dec 2025 | |
| $0.10 | $0.10 | $0.010 | — | 2 Dec 2025 | |
| $0.15 | $0.15 | $0.015 | — | 2 Dec 2025 | |
Mistral Medium 3.5Shuts down 30 Apr |
$1.50 | $7.50 | $0.15 | 256K | 28 Apr 2026 |
Mistral Large 3Shuts down 30 Apr |
$0.50 | $1.50 | $0.050 | 256K | 2 Dec 2025 |
Version details
Codestral 25.08
$0.30 / $0.90 per 1M tokenscached $0.030 · batch 50% off · 256K context
- A code-specialised model; excluded from general comparisons and the index.
Checked 2 Nov 2026 against Mistral pricing .
Mistral Small 4
$0.15 / $0.60 per 1M tokenscached $0.015 · batch 50% off · 256K context
Checked 2 Nov 2026 against Mistral pricing .
Ministral 3 14B
$0.20 / $0.20 per 1M tokenscached $0.020 · batch 50% off
Checked 2 Nov 2026 against Mistral pricing .
Ministral 3 3B
$0.10 / $0.10 per 1M tokenscached $0.010 · batch 50% off
Checked 2 Nov 2026 against Mistral pricing .
Ministral 3 8B
$0.15 / $0.15 per 1M tokenscached $0.015 · batch 50% off
Checked 2 Nov 2026 against Mistral pricing .
Mistral Medium 3.5
Shuts down 30 Apr$1.50 / $7.50 per 1M tokenscached $0.15 · batch 50% off · 256K context
- Deprecated on 31 October 2026 and no longer on Mistral's price list; retires on 30 April 2027.
Newer option: Mistral Large 4 at $0.70 / $2.10. Checked 2 Nov 2026 against Mistral pricing , Mistral model card and Mistral models overview .
Mistral Large 3
Shuts down 30 Apr$0.50 / $1.50 per 1M tokenscached $0.050 · batch 50% off · 256K context
- Deprecated on 31 October 2026 and no longer on Mistral's price list; retires on 30 April 2027.
Newer option: Mistral Large 4 at $0.70 / $2.10. Checked 2 Nov 2026 against Mistral pricing , Mistral model card and Mistral models overview .
Compare Mistral Large 4
No ready-made comparisons for this family yet.
Price history
All changes- Mistral deprecated Large 3 and Medium 3.5, which retire on 30 April 2027, and pointed users to Mistral Large 4, now listed at $0.70 / $2.10 per 1M tokens (it was $1.36 / $4.18 when we checked on 22 October). Source
- Mistral retired Nemo, Small 3.2, Devstral 2 and Magistral 1.2. Source
- Mistral launched Medium 3.5 at $1.50 / $7.50. Source
Coming up
- Mistral retires Large 3 and Medium 3.5. It recommends moving to Mistral Large 4. Source
Every upcoming shutdown and price change
Every dated price: price records · CSV. Cite as: AI Token Price, Mistral pricing, https://aitokenprice.com/pricing/mistral#price-history
Mistral AI pricing rules
- Batch requests cost 50% less.
- Nemo, Small 3.2, Devstral 2 and Magistral 1.2 were retired on 31 July 2026.
Retired Mistral models
- Retired 31 Jul 2026Mistral Small 3.2. Use Mistral Small 4 instead.
- Retired 31 Jul 2026Mistral Nemo. Use Ministral 3 8B instead.
Questions
How much does Mistral Large 4 cost?
As of 2 Nov 2026, Mistral Large 4 costs $0.70 per million input tokens and $2.10 per million output tokens, and $0.07 per million cached input tokens. Batch requests cost 50% less. Prices checked 2 Nov 2026.
How much does Codestral 25.08 cost?
Codestral 25.08 costs $0.30 per million input tokens and $0.90 per million output tokens, with cached input at $0.03. Checked 2 Nov 2026.
How much does Mistral Small 4 cost?
Mistral Small 4 costs $0.15 per million input tokens and $0.60 per million output tokens, with cached input at $0.015. Checked 2 Nov 2026.
How much does Ministral 3 14B cost?
Ministral 3 14B costs $0.20 per million input tokens and $0.20 per million output tokens, with cached input at $0.02. Checked 2 Nov 2026.
What is the cheapest Mistral model?
Ministral 3 3B, at $0.10 / $0.10 per million input / output tokens.
What is Mistral Large 4's context window?
1M tokens.
Does Mistral Large 4 support prompt caching?
Yes. Repeated input read from the cache costs $0.07 per million tokens, 90% less than normal input.