Compare / Claude Haiku 4.5 vs Mistral Small 3.1 24B
Claude Haiku 4.5 vs Mistral Small 3.1 24B
Which is cheaper, which is more capable, and which does better at each kind of work, from 21 results on 3 sources that measured both models.
Claude Haiku 4.5 or Mistral Small 3.1 24B? The short answer
- Mistral Small 3.1 24B is cheaper: $0.40 against $2.00 per million tokens, 5.0 times less.
- Claude Haiku 4.5 scores higher on the Artificial Analysis Intelligence Index: 16.9 against 7.1.
- On our business benchmarks, Claude Haiku 4.5 ranks higher for agents and tool use, reasoning and knowledge and writing.
At a glance
| Claude Haiku 4.5 | Mistral Small 3.1 24B | |
|---|---|---|
| Price per million tokens | $1.00 in · $5.00 out | $0.35 in · $0.56 out |
| Blended (3 in : 1 out) | $2.00 | $0.40 |
| Cheapest host, blended | $2.00 (Azure) | – |
| Intelligence Index | 16.9 | 7.1 |
| Spring Prompt overall | 8 | – |
| Output speed, tokens a second | 102 | – |
| Context window | 200,000 tokens | 128,000 tokens |
| Weights | Closed | Open |
| Developer based in | the US | France |
Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.
Which is better for what
Each model's rank among models available today, for each kind of work, from our own benchmarks and licensed sources. The better rank is in green.
| Use case | Claude Haiku 4.5 | Mistral Small 3.1 24B |
|---|---|---|
| Product listingsTurning a sparse product feed and photos into listings that can go live | 18 of 19 | not measured |
| Decks from an analysisTurning a finished analysis into a deck you could present as it is | 16 of 18 | not measured |
| Marketing planningPlanning a year of ad spend without overspending | 17 of 18 | not measured |
| Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls | 88 of 161 | 114 of 161 |
| Reasoning and knowledgeHard questions across science, maths and general knowledge | 123 of 195 | 172 of 195 |
| WritingWhat people prefer in blind comparisons, and judged writing quality | 84 of 187 | 157 of 187 |
| Sticking to the factsSummarising without inventing things, and factual answers | 118 of 153 | not measured |
| Speed under pressureGood decisions against a real clock (fast chess) | 16 of 21 | not measured |
Every shared result
| Benchmark | Metric | Claude Haiku 4.5 | Mistral Small 3.1 24B |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,420 | 1,297 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,387 | 1,270 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,452 | 1,302 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,415 | 1,294 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,414 | 1,303 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,399 | 1,290 |
| Artificial Analysisreported | Artificial Analysis Coding Indexindex score | 43.9 | 26.3 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 16.9 | 7.10 |
| Artificial Analysisreported | GPQA Diamond% of questions | 67.2% | 45.4% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 10.4% | 4.3% |
| Artificial Analysisreported | IFBench% of instructions | 54.3% | 29.9% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 74.3% | 22.3% |
| Artificial Analysisreported | SciCode% of problems | 42.2% | 27.8% |
| Artificial Analysisreported | Terminal-Bench 2.1% of tasks | 44.2% | 26.2% |
| Artificial Analysisreported | Terminal-Bench 4.0% of tasks | 0.0% | 0.0% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 27.3% | 7.6% |
| Artificial Analysisreported | Τ-bench banking% of tasks | 9.3% | 7.4% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 54.7% | 25.1% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 6.0% | 11.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.39 | 0.35 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 52.2 | 32.1 |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.
Quick answers
Is Claude Haiku 4.5 better than Mistral Small 3.1 24B?
It depends on the task. Among models available today, on our benchmarks and licensed sources, Claude Haiku 4.5 ranks higher for agents and tool use, reasoning and knowledge and writing.
Which is cheaper, Claude Haiku 4.5 or Mistral Small 3.1 24B?
Mistral Small 3.1 24B costs $0.40 per million tokens (three input to one output), against $2.00 for Claude Haiku 4.5.
Run Claude Haiku 4.5 and Mistral Small 3.1 24B on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.