Compare
Four models, one page
Pick the models you are actually choosing between. Every row is scaled within itself, so a bar means something next to its neighbours and nothing across rows.
| Specification | V4 Flash 0731DeepSeek | Claude Opus 5Anthropic |
|---|---|---|
| Released | Jul 31, 2026 | Jul 24, 2026 |
| Context window | 1.05M | 1M |
| Max output | 66K | 128K |
| Input / 1M | $0.09 | $5 |
| Output / 1M | $0.18 | $25 |
| Cost per task | $0.02 | $2.25 |
| Arena Elo | 1,279 | 1,393 |
| Serving providers | 23 | 5 |
| Parameters | 304.2B | Undisclosed |
| Licence | mit | Proprietary |
| Capabilities |
|
|
Metrics side by side
Each row is scaled to the largest value in that row — bars compare within a row, never across rows
- V4 Flash 0731
- Claude Opus 5
Intelligence Index
V4 Flash 073151.8Claude Opus 563.1Coding Index
V4 Flash 073169.1Claude Opus 578.0Agentic Index
V4 Flash 073148.4Claude Opus 559.2Output speed
V4 Flash 0731148 t/sClaude Opus 587 t/sContext window
V4 Flash 07311.05MClaude Opus 51MLatency · lower is better
V4 Flash 0731324msClaude Opus 51.12sBlended price / 1M · lower is better
V4 Flash 0731$0.113Claude Opus 5$10Cost per task · lower is better
V4 Flash 0731$0.02Claude Opus 5$2.25
Rows marked “lower is better” still draw a longer bar for a larger number — read the value, not just the length. Arena Elo is in the specification table above instead: it has no meaningful zero, so a bar would flatten the gaps.
View as table
| Metric | V4 Flash 0731 | Claude Opus 5 |
|---|---|---|
| Intelligence Index | 51.8 | 63.1 |
| Coding Index | 69.1 | 78.0 |
| Agentic Index | 48.4 | 59.2 |
| Output speed | 148 t/s | 87 t/s |
| Context window | 1.05M | 1M |
| Latency · lower is better | 324ms | 1.12s |
| Blended price / 1M · lower is better | $0.113 | $10 |
| Cost per task · lower is better | $0.02 | $2.25 |
Evaluation scores
Percentage correct on a common 0–100% scale
- V4 Flash 0731
- Claude Opus 5
GPQA Diamond
V4 Flash 073190.8%Claude Opus 593.2%Humanity's Last Exam
V4 Flash 073138.6%Claude Opus 554.9%SciCode
V4 Flash 073149.9%Claude Opus 555.7%τ²-bench
V4 Flash 073139.4%Claude Opus 542.1%AA-LCR long context
V4 Flash 073174.3%Claude Opus 575.7%
A missing bar means that evaluation was not run for that model — it is not a zero.
View as table
| Evaluation | V4 Flash 0731 | Claude Opus 5 |
|---|---|---|
| GPQA Diamond | 90.8% | 93.2% |
| Humanity's Last Exam | 38.6% | 54.9% |
| SciCode | 49.9% | 55.7% |
| τ²-bench | 39.4% | 42.1% |
| Terminal-Bench Hard | — | — |
| LiveCodeBench | — | — |
| AA-LCR long context | 74.3% | 75.7% |
| AIME 2025 | — | — |
Estimated from list pricing: 50K input tokens plus 80K output tokens for reasoning models (25K for non-reasoning).