Loading benchmark catalogue
Compare performance benchmarks, coding intelligence, reasoning evaluations, context limits, and API economics between Claude Haiku 5.5 (Anthropic) and DeepSeek V4.1 Flash 1017 (DeepSeek).
| Metric / Capability | Claude Haiku 5.5 | DeepSeek V4.1 Flash 1017 |
|---|---|---|
| Developer / Organization | Anthropic | DeepSeek |
| Release Date | 2026-10-07 | 2026-10-17 |
| Deployment / Weights | Proprietary Cloud API | Open Weights |
| Context Window | 1M tokens (1,000,000 tokens) | 1.0M tokens (1,048,576 tokens) |
| Input Pricing (/1M Tokens) | $0.1 | Free (Open Weights) |
| Output Pricing (/1M Tokens) | $0.5 | Free (Open Weights) |
| Benchmark / Domain | Claude Haiku 5.5 | DeepSeek V4.1 Flash 1017 | Advantage |
|---|---|---|---|
Artificial Analysis Intelligence Index Composite index aggregating 10+ evaluations of general reasoning and knowledge. | 43.4% | 39.5% | Claude Haiku 5.5 (+10%) |
Artificial Analysis Coding Agent Index v2.0 Composite index of coding-agent systems (model + harness) across FrontierCode v1.1, SWE-Atlas-QnA, and Terminal-Bench v4. | 37.3% | 46.3% | DeepSeek V4.1 Flash 1017 (+24%) |
Terminal-Bench 4.0 Sandboxed command-line tasks requiring shell and file-system operation. | 39.2% | 42.3% | DeepSeek V4.1 Flash 1017 (+8%) |
ARC-AGI-2 Novel visual grid puzzles requiring abstract rule induction. | 61.3% | — | — |
GPQA Diamond Graduate-level physics, chemistry, and biology questions written by domain experts. | — | 90.9% | — |
Humanity's Last Exam Expert-written questions across hundreds of disciplines designed to resist saturation. | 45.9% | 42.6% | Claude Haiku 5.5 (+8%) |
FrontierMath v2 (Tiers 1-3) Unpublished, research-level mathematics problems across Tiers 1-3. | 75.1% | 67.4% | Claude Haiku 5.5 (+11%) |
Design Arena ELO Human-preference ELO across UI, 3D, data-visualization, and game-dev generations. | 1344 | 1362 | DeepSeek V4.1 Flash 1017 (+1%) |
Both models perform neck-and-neck across available benchmarks, with Claude Haiku 5.5 and DeepSeek V4.1 Flash 1017 each excelling in distinct tasks.
DeepSeek V4.1 Flash 1017 has the larger context capacity of 1.0M tokens (1,048,576 tokens), compared to 1M tokens for Claude Haiku 5.5.
DeepSeek V4.1 Flash 1017 is more economical. Input tokens cost $Free (Open Weights)/M tokens compared to $0.1/M tokens for Claude Haiku 5.5.
Claude Haiku 5.5 is a proprietary cloud API. DeepSeek V4.1 Flash 1017 is an open-weights model.
Analyze 130+ models side-by-side using interactive radar charts, cost-efficiency quadrant plots, and live dataset filters on the full NeuBench index.
Open Interactive Comparator