vs
13:5
overall capability wins (0 ties across 18 scored). Claude Opus 5 leads overall.
- Reasoning4:2
- Coding4:1
- Arena1:0
- Agents2:0
- Tool use0:1
- Math1:0
- Composite indices2:0
- Pricing0:3
- Speed2:0
- Specs1:1
Verdict
Claude Opus 5 leads coding (widest gap: +12.7 on WebDev Elo); Claude Opus 5 ranks higher on LMArena (+8 Elo); Kimi K3 offers larger context (1.0M vs 1M).
- Claude Opus 5 leads coding (widest gap: +12.7 on WebDev Elo)
- Claude Opus 5 ranks higher on LMArena (+8 Elo)
- Kimi K3 offers larger context (1.0M vs 1M)
- Claude Opus 5 allows ~2.0x more max output (128K vs 64K)
- Kimi K3 ships open weights (Kimi K3 License)
- Kimi K3 is ~1.7x cheaper on a blended token basis than Claude Opus 5
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Claude Opus 5
- $11.25
- Kimi K3
- $6.75
Kimi K3 is about 1.7x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
Anthropic
Moonshot
Family
Claude Opus
Kimi
License
Proprietary
Kimi K3 License
Open weights
No
Yes
Release
Jul 24, 2026
Jul 16, 2026
Knowledge cutoff
-
-
API / provider
Anthropic
Moonshot
Modalities
text, image → text
text, image, video → text
Specs
Context window
1M
1.0M
Max output
128K
64K
Parameters
—
2.8T (104B act.)
Pricing
Input $/1M
$5
$3
Output $/1M
$25
$15
Blended $/1M (3∶1)
$10
$6
Speed
tok/s
72
60
TTFT (s)
0.55
0.6
Reasoning
MMLU-Pro
—
—
GPQA Diamond
93.2
93.5
Humanity's Last Exam
54.9
46.9
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
60.9
47.6
AA-LCR v1.1
79.3
88.7
CritPt
29.1
23.4
MMMU-Pro
84.7
80.5
IFBench
—
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
89.1
85.0
Aider Polyglot
—
—
Terminal-Bench 3
42.7
—
BigCodeBench
—
—
SciCode
56.4
59.5
CursorBench
70.0
60.8
SWE-Rebench
63.4
—
NL2Repo-Bench
—
—
DeepSWE
73.7
69.0
WebDev Arena
1,687
1,674
Terminal-Bench 4.0
51.8
—
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
1,493
1,485
| Metric | Claude Opus 5 | Kimi K3 | Delta |
|---|---|---|---|
| Identity | |||
| Organization | Anthropic | Moonshot | — |
| Family | Claude Opus | Kimi | — |
| License | Proprietary | Kimi K3 License | — |
| Open weights | No | Yes | — |
| Release | Jul 24, 2026 | Jul 16, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | Anthropic | Moonshot | — |
| Modalities | text, image → text | text, image, video → text | — |
| Specs | |||
| Context window | 1M | 1.0M | B +49K |
| Max output | 128K | 64K | A +64K |
| Parameters | — | 2.8T (104B act.) | — |
| Pricing | |||
| Input $/1M | $5 | $3 | B +$2 |
| Output $/1M | $25 | $15 | B +$10 |
| Blended $/1M (3∶1) | $10 | $6 | B +$4 |
| Speed | |||
| tok/s | 72 | 60 | A +12 |
| TTFT (s) | 0.55 | 0.6 | A +0.05 |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | 93.2 | 93.5 | B +0.3 pts |
| Humanity's Last Exam | 54.9 | 46.9 | A +8.0 pts |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | 60.9 | 47.6 | A +13.3 pts |
| AA-LCR v1.1 | 79.3 | 88.7 | B +9.4 pts |
| CritPt | 29.1 | 23.4 | A +5.7 pts |
| MMMU-Pro | 84.7 | 80.5 | A +4.2 pts |
| IFBench | — | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | 89.1 | 85.0 | A +4.1 pts |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | 42.7 | — | — |
| BigCodeBench | — | — | — |
| SciCode | 56.4 | 59.5 | B +3.1 pts |
| CursorBench | 70.0 | 60.8 | A +9.2 pts |
| SWE-Rebench | 63.4 | — | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | 73.7 | 69.0 | A +4.7 pts |
| WebDev Arena | 1,687 | 1,674 | A +13 Elo |
| Terminal-Bench 4.0 | 51.8 | — | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | 1,493 | 1,485 | A +8 Elo |
Claude Opus 5
- Context
- 1M / 128K out
- Parameters
- —
- Price
- $5 / $25
- Speed
- 72 tok/s · 0.55s TTFT
- Modalities
- text, image → text
- License
- Proprietary
Anthropic's July 2026 flagship - top Artificial Analysis Intelligence Index and strong agentic coding at Opus 4.8 pricing.
Kimi K3
- Context
- 1.0M / 64K out
- Parameters
- 2.8T (104B act.)
- Price
- $3 / $15
- Speed
- 60 tok/s · 0.6s TTFT
- Modalities
- text, image, video → text
- License
- Kimi K3 License
Largest open-weight model (2.8T MoE / 104B active) — official card reports strong reasoning and agentic coding results.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.