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 | Kimi K3Moonshot AI | Claude Opus 5Anthropic |
|---|---|---|
| Released | Jul 16, 2026 | Jul 24, 2026 |
| Context window | 1.05M | 1M |
| Max output | — | 128K |
| Input / 1M | $3 | $5 |
| Output / 1M | $15 | $25 |
| Cost per task | $1.35 | $2.25 |
| Arena Elo | 1,455 | 1,393 |
| Serving providers | 9 | 5 |
| Parameters | 2779.9B | Undisclosed |
| Licence | other | 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
- Kimi K3
- Claude Opus 5
Intelligence Index
Kimi K359.7Claude Opus 563.1Coding Index
Kimi K376.2Claude Opus 578.0Agentic Index
Kimi K354.3Claude Opus 559.2Output speed
Kimi K371 t/sClaude Opus 587 t/sContext window
Kimi K31.05MClaude Opus 51MLatency · lower is better
Kimi K31.15sClaude Opus 51.12sBlended price / 1M · lower is better
Kimi K3$6Claude Opus 5$10Cost per task · lower is better
Kimi K3$1.35Claude 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 | Kimi K3 | Claude Opus 5 |
|---|---|---|
| Intelligence Index | 59.7 | 63.1 |
| Coding Index | 76.2 | 78.0 |
| Agentic Index | 54.3 | 59.2 |
| Output speed | 71 t/s | 87 t/s |
| Context window | 1.05M | 1M |
| Latency · lower is better | 1.15s | 1.12s |
| Blended price / 1M · lower is better | $6 | $10 |
| Cost per task · lower is better | $1.35 | $2.25 |
Evaluation scores
Percentage correct on a common 0–100% scale
- Kimi K3
- Claude Opus 5
GPQA Diamond
Kimi K393.5%Claude Opus 593.2%Humanity's Last Exam
Kimi K346.9%Claude Opus 554.9%SciCode
Kimi K358.7%Claude Opus 555.7%τ²-bench
Kimi K346.0%Claude Opus 542.1%AA-LCR long context
Kimi K382.7%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 | Kimi K3 | Claude Opus 5 |
|---|---|---|
| GPQA Diamond | 93.5% | 93.2% |
| Humanity's Last Exam | 46.9% | 54.9% |
| SciCode | 58.7% | 55.7% |
| τ²-bench | 46.0% | 42.1% |
| Terminal-Bench Hard | — | — |
| LiveCodeBench | — | — |
| AA-LCR long context | 82.7% | 75.7% |
| AIME 2025 | — | — |
Estimated from list pricing: 50K input tokens plus 80K output tokens for reasoning models (25K for non-reasoning).