Model comparison
Claude Haiku 4.5 vs DeepSeek-V3.2-Speciale
Claude Haiku 4.5 and DeepSeek-V3.2-Speciale score almost the same on the Noometry Index (39.5 vs 39.7), so choose on price, context window or the category you care about most.
Last verified . 1 shared benchmarks.
DeepSeekSummary
- They share 1 benchmark with published results for both. Claude Haiku 4.5 scores higher in 2 categories and DeepSeek-V3.2-Speciale in 1 category; 3 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek-V3.2-Speciale leads 32.9 to 15.1.
- DeepSeek-V3.2-Speciale is cheaper at $0.58 / $1.68 per million input/output tokens, against $1 / $5 for Claude Haiku 4.5.
- Claude Haiku 4.5 accepts more context: 200K tokens versus 128K.
- DeepSeek-V3.2-Speciale has downloadable open weights; the other is API-only.
Side by side
| Claude Haiku 4.5 | DeepSeek-V3.2-Speciale | |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 39.5 | 39.7 |
| Released | 2025-10-15 | 2025-12-01 |
| Weights | Proprietary | Open |
| Context window | 200K | 128K |
| Max output | 64K | 128K |
| Input $ / M tokens | $1 | $0.58 |
| Output $ / M tokens | $5 | $1.68 |
| Results tracked | 53 | 3 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Haiku 4.5 leads
Claude Haiku 4.5: 44.0 (#78), DeepSeek-V3.2-Speciale: 40.4 (#140)
| Benchmark | Claude Haiku 4.5 | DeepSeek-V3.2-Speciale |
|---|---|---|
| WeirdML | 45.4% | 46.7% |
| SWE-bench Verified (bash only) | 66.6% | — |
| LMArena WebDev | 1330 | — |
| SWE-bench Multilingual | 64.7% | — |
| SciCode | 43.3% | — |
| LMArena Coding | 1453 | — |
| ALE-Bench | 653.48 | — |
Agentic & Tool Use Not comparable
Claude Haiku 4.5: 33.6 (#52), DeepSeek-V3.2-Speciale: —
| Benchmark | Claude Haiku 4.5 | DeepSeek-V3.2-Speciale |
|---|---|---|
| Terminal-Bench | 35.5% | — |
| Berkeley Function Calling Leaderboard | 68.7% | — |
| DeepResearch Bench | 45.5% | — |
| BALROG | 31.2% | — |
| ExploitBench | 13.7% | — |
| Vending-Bench 2 | 458.89 | — |
Reasoning DeepSeek-V3.2-Speciale leads
Claude Haiku 4.5: 15.1 (#320), DeepSeek-V3.2-Speciale: 32.9 (#73)
| Benchmark | Claude Haiku 4.5 | DeepSeek-V3.2-Speciale |
|---|---|---|
| ARC-AGI-2 | 4% | — |
| SimpleBench | — | 52.6% |
| NYT Connections (extended) | 14.3% | — |
| ARC-AGI-1 | 47.7% | — |
| CritPt | 0% | — |
| Chess Puzzles | 8% | — |
| LMArena Hard Prompts | 1420 | — |
| DTBench | 73.6% | — |
| LMCA | 30.9% | — |
| Epoch Capabilities Index | 142.41 | — |
| ForecastBench | 61.4 | — |
Math Not comparable
Claude Haiku 4.5: 44.9 (#78), DeepSeek-V3.2-Speciale: —
| Benchmark | Claude Haiku 4.5 | DeepSeek-V3.2-Speciale |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.7% | — |
| Omni-MATH | 56.1% | — |
| LMArena Math | 1396 | — |
| MATH Level 5 | 96.4% | — |
| FrontierMath (Feb 2025 set) | 5.9% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Not comparable
Claude Haiku 4.5: 37.7 (#153), DeepSeek-V3.2-Speciale: —
| Benchmark | Claude Haiku 4.5 | DeepSeek-V3.2-Speciale |
|---|---|---|
| GPQA Diamond | 71.2% | — |
| SimpleQA Verified | 13.2% | — |
| MMLU-Pro | 77.7% | — |
| Vectara Hallucination Rate | 9.8% | — |
| GPQA (HELM) | 60.5% | — |
| LMArena Expert | 1442 | — |
Multimodal Not comparable
Claude Haiku 4.5: 26.8 (#118), DeepSeek-V3.2-Speciale: —
| Benchmark | Claude Haiku 4.5 | DeepSeek-V3.2-Speciale |
|---|---|---|
| Blueprint-Bench 2 | 0% | — |
| LMArena Document | 1420 | — |
Multilingual Not comparable
Claude Haiku 4.5: 49.9 (#129), DeepSeek-V3.2-Speciale: —
| Benchmark | Claude Haiku 4.5 | DeepSeek-V3.2-Speciale |
|---|---|---|
| LMArena Non-English | 1377 | — |
| LMArena Chinese | 1417 | — |
| LMArena French | 1408 | — |
| LMArena German | 1375 | — |
| LMArena Japanese | 1339 | — |
| LMArena Korean | 1347 | — |
| LMArena Russian | 1381 | — |
| LMArena Spanish | 1420 | — |
Instruction Following Not comparable
Claude Haiku 4.5: 71.4 (#149), DeepSeek-V3.2-Speciale: —
| Benchmark | Claude Haiku 4.5 | DeepSeek-V3.2-Speciale |
|---|---|---|
| IFEval | 80.1% | — |
| LMArena Instruction Following | 1414 | — |
Long Context Not comparable
Claude Haiku 4.5: 43.6 (#92), DeepSeek-V3.2-Speciale: —
| Benchmark | Claude Haiku 4.5 | DeepSeek-V3.2-Speciale |
|---|---|---|
| LMArena Longer Query | 1427 | — |
Writing & Preference Claude Haiku 4.5 leads
Claude Haiku 4.5: 57.9 (#123), DeepSeek-V3.2-Speciale: 46.0 (#222)
| Benchmark | Claude Haiku 4.5 | DeepSeek-V3.2-Speciale |
|---|---|---|
| LMArena Text | 1396 | — |
| LMArena Creative Writing | 1372 | — |
| EQ-Bench Creative Writing | — | 1276 |
| WildBench | 83.9% | — |
| EQ-Bench 4 | 1064 | — |
| LMArena Multi-Turn | 1409 | — |
Frequently asked questions
Is Claude Haiku 4.5 better than DeepSeek-V3.2-Speciale?
Claude Haiku 4.5 and DeepSeek-V3.2-Speciale score almost the same on the Noometry Index (39.5 vs 39.7), so choose on price, context window or the category you care about most.
Which is cheaper, Claude Haiku 4.5 or DeepSeek-V3.2-Speciale?
DeepSeek-V3.2-Speciale is cheaper. It lists at $0.58 per million input tokens and $1.68 per million output tokens; Claude Haiku 4.5 lists at $1 and $5.
Is Claude Haiku 4.5 or DeepSeek-V3.2-Speciale better for coding?
Claude Haiku 4.5 scores higher on coding benchmarks: 44.0 versus 40.4 in the Noometry coding category.
Which has the bigger context window?
Claude Haiku 4.5 does, with 200K tokens against 128K.
How many benchmarks do Claude Haiku 4.5 and DeepSeek-V3.2-Speciale share?
1 benchmark has published results for both models. Claude Haiku 4.5 has 53 scored results on Noometry and DeepSeek-V3.2-Speciale has 3.