Loading benchmark catalogue
Compare performance benchmarks, coding intelligence, reasoning evaluations, context limits, and API economics between DeepSeek V4 Pro 0813 (DeepSeek) and Claude Opus 5 (Anthropic).
| Metric / Capability | DeepSeek V4 Pro 0813 | Claude Opus 5 |
|---|---|---|
| Developer / Organization | DeepSeek | Anthropic |
| Release Date | 2026-08-12 | 2026-07-24 |
| Deployment / Weights | Open Weights | Proprietary Cloud API |
| Context Window | 1.0M tokens (1,048,576 tokens) | 1M tokens (1,000,000 tokens) |
| Input Pricing (/1M Tokens) | Free (Open Weights) | $5 |
| Output Pricing (/1M Tokens) | Free (Open Weights) | $25 |
| Benchmark / Domain | DeepSeek V4 Pro 0813 | Claude Opus 5 | Advantage |
|---|---|---|---|
Artificial Analysis Intelligence Index Composite index aggregating 10+ evaluations of general reasoning and knowledge. | 36.0% | 50.8% | Claude Opus 5 (+41%) |
Artificial Analysis Coding Index Composite index of coding and software-engineering evaluations. | 68.8% | 78.0% | Claude Opus 5 (+13%) |
Artificial Analysis Agentic Index Composite index of tool-use and long-horizon agentic tasks. | 41.3% | 56.5% | Claude Opus 5 (+37%) |
SWE-bench Pro Harder, contamination-resistant successor to SWE-bench spanning multiple languages. | 55.4% | 79.2% | Claude Opus 5 (+43%) |
Terminal-Bench 4.0 Sandboxed command-line tasks requiring shell and file-system operation. | — | 53.9% | — |
ARC-AGI-2 Novel visual grid puzzles requiring abstract rule induction. | 61.3% | 90.4% | Claude Opus 5 (+47%) |
GPQA Diamond Graduate-level physics, chemistry, and biology questions written by domain experts. | 90.1% | — | — |
Humanity's Last Exam Expert-written questions across hundreds of disciplines designed to resist saturation. | 42.7% | 56.3% | Claude Opus 5 (+32%) |
FrontierMath v2 (Tiers 1-3) Unpublished, research-level mathematics problems across Tiers 1-3. | 45.3% | 85.6% | Claude Opus 5 (+89%) |
Design Arena ELO Human-preference ELO across UI, 3D, data-visualization, and game-dev generations. | — | 1358 | — |
Claude Opus 5 outperforms DeepSeek V4 Pro 0813 across 7 evaluated benchmarks compared to 0 for DeepSeek V4 Pro 0813, excelling in Artificial Analysis Intelligence Index, Artificial Analysis Coding Index, Artificial Analysis Agentic Index.
DeepSeek V4 Pro 0813 offers a larger context capacity of 1.0M tokens (1,048,576 tokens), compared to 1M tokens for Claude Opus 5.
DeepSeek V4 Pro 0813 is more cost-effective. Input tokens cost $Free (Open Weights)/M tokens vs $5/M tokens for Claude Opus 5.
DeepSeek V4 Pro 0813 is an open-weights model. Claude Opus 5 is a proprietary cloud API.
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