Mistral: Models Intelligence, Performance & Price
Analysis of Mistral's models across key metrics including quality, price, output speed, latency, context window & more. This analysis is intended to support you in choosing the best model provided by Mistral for your use-case. For more details including relating to our methodology, see our FAQs. Models analyzed: Pixtral Large, Mistral Large 2 (Nov '24), Mistral Large 2 (Jul '24), Mistral Small 3, Mistral Small (Sep '24), Mixtral 8x22B, Pixtral 12B, Ministral 8B, Mistral NeMo, Ministral 3B, Mixtral 8x7B, Codestral-Mamba, Mistral Small 3.1, Mistral Saba, Codestral (Jan '25), Mistral Small (Feb '24), Mistral Large (Feb '24), Mistral 7B, Mistral Medium, and Codestral (May '24).
Link:
Mistral Model Comparison Summary
Intelligence: Mistral Large 2 (Nov '24) and Pixtral Large are the highest quality models offered by Mistral, followed by Mistral Large 2 (Jul '24), Mistral Small 3.1 & Mistral Small 3.Output Speed (tokens/s): Ministral 3B (218 t/s) and Codestral (Jan '25) (183 t/s) are the fastest models offered by Mistral, followed by Mistral Small 3.1, Mistral Small (Feb '24) & Ministral 8B.Latency (seconds): Ministral 3B (0.32s) and Mistral Small (Feb '24) (0.33s) are the lowest latency models offered by Mistral, followed by Mistral Small 3.1, Codestral (Jan '25) & Pixtral 12B.Blended Price ($/M tokens): Ministral 3B ($0.04) and Ministral 8B ($0.10) are the cheapest models offered by Mistral, followed by Mistral Small 3, Pixtral 12B & Mistral NeMo.Context Window Size: Codestral-Mamba (256k) and Codestral (Jan '25) (256k) are the largest context window models offered by Mistral, followed by Pixtral Large, Mistral Large 2 (Nov '24) & Mistral Large 2 (Jul '24).
Mistral Large 2 (Nov '24) and Pixtral Large are the highest quality models offered by Mistral, followed by Mistral Large 2 (Jul '24), Mistral Small 3.1 & Mistral Small 3.Output Speed (tokens/s): Ministral 3B (218 t/s) and Codestral (Jan '25) (183 t/s) are the fastest models offered by Mistral, followed by Mistral Small 3.1, Mistral Small (Feb '24) & Ministral 8B.Latency (seconds): Ministral 3B (0.32s) and Mistral Small (Feb '24) (0.33s) are the lowest latency models offered by Mistral, followed by Mistral Small 3.1, Codestral (Jan '25) & Pixtral 12B.Blended Price ($/M tokens): Ministral 3B ($0.04) and Ministral 8B ($0.10) are the cheapest models offered by Mistral, followed by Mistral Small 3, Pixtral 12B & Mistral NeMo.Context Window Size: Codestral-Mamba (256k) and Codestral (Jan '25) (256k) are the largest context window models offered by Mistral, followed by Pixtral Large, Mistral Large 2 (Nov '24) & Mistral Large 2 (Jul '24).Highlights
Intelligence
Artificial Analysis Intelligence Index; Higher is better
Speed
Output Tokens per Second; Higher is better
Price
USD per 1M Tokens; Lower is better
Parallel Queries:
Prompt Length:
Features | Model Intelligence | Price | Output tokens/s | Latency | End-to-End Response Time | ||||
|---|---|---|---|---|---|---|---|---|---|
Further Analysis | |||||||||
 | Mistral Large 2 (Nov '24) | 128k | 38 | $3.00 | 37.1 | 0.58 | 14.06 | N/A | |
| Pixtral Large | 128k | 37 | $3.00 | 32.9 | 0.49 | 15.69 | N/A | |
| Mistral Large 2 (Jul '24) | 128k | 37 | $3.00 | 36.0 | 0.52 | 14.41 | N/A | |
| Mistral Small 3.1 | 128k | 35 | $0.15 | 166.6 | 0.34 | 3.34 | N/A | |
| Mistral Small 3 | 32k | 35 | $0.15 | 120.7 | 0.41 | 4.56 | N/A | |
| Codestral (Jan '25) | 256k | 28 | $0.45 | 183.1 | 0.34 | 3.07 | N/A | |
| Mistral Small (Sep '24) | 33k | 27 | $0.30 | 59.9 | 0.55 | 8.89 | N/A | |
| Mistral Large (Feb '24) | 33k | 26 | $6.00 | 31.3 | 0.56 | 16.54 | N/A | |
| Mixtral 8x22B | 65k | 26 | $3.00 | 58.3 | 0.40 | 8.98 | N/A | |
| Pixtral 12B | 128k | 23 | $0.15 | 104.9 | 0.35 | 5.12 | N/A | |
| Mistral Small (Feb '24) | 33k | 23 | $1.50 | 160.3 | 0.33 | 3.45 | N/A | |
| Mistral Medium | 33k | 23 | $4.09 | 40.2 | 0.46 | 12.90 | N/A | |
| Ministral 8B | 128k | 22 | $0.10 | 144.7 | 0.38 | 3.83 | N/A | |
| Codestral (May '24) | 33k | 20 | $0.30 | 96.5 | 0.42 | 5.61 | N/A | |
| Ministral 3B | 128k | 20 | $0.04 | 217.7 | 0.32 | 2.61 | N/A | |
| Mistral NeMo | 128k | 20 | $0.15 | 125.4 | 0.36 | 4.35 | N/A | |
| Mixtral 8x7B | 33k | 17 | $0.70 | 68.9 | 0.39 | 7.65 | N/A | |
| Codestral-Mamba | 256k | 14 | $0.25 | 94.3 | 0.52 | 5.82 | N/A | |
| Mistral 7B | 8k | 10 | $0.25 | 102.3 | 0.45 | 5.34 | N/A | |
| Mistral Saba | 32k | $0.30 | 95.8 | 0.43 | 5.65 | N/A | ||
Key definitions
Context window: Maximum number of combined input & output tokens. Output tokens commonly have a significantly lower limit (varied by model).
Output Speed: Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).
Latency (Time to First Token): Time to first token received, in seconds, after API request sent. For reasoning models which share reasoning tokens, this will be the first reasoning token. For models which do not support streaming, this represents time to receive the completion.
Price: Price per token, represented as USD per million Tokens. Price is a blend of Input & Output token prices (3:1 ratio).
Output Price: Price per token generated by the model (received from the API), represented as USD per million Tokens.
Input Price: Price per token included in the request/message sent to the API, represented as USD per million Tokens.
Time period: Metrics are 'live' and are based on the past 72 hours of measurements, measurements are taken 8 times a day for single requests and 2 times per day for parallel requests.