Head to head
Kimi K3 vs Claude Opus 5
Kimi K3 from Moonshot AI against Claude Opus 5 from Anthropic — specification, price and every benchmark both makers have published, in one table.
Benchmarks
No shared benchmarks
These two models have no benchmark in common with published figures for both, so there is nothing to compare directly. The specification and price rows below are still like for like.
Price
Kimi K3 is cheaper
On a 3:1 input-to-output mix, Kimi K3 costs $6.00 per million tokens against $10.00 for Claude Opus 5 — about 1.7× less. Remember that a reasoning model bills its thinking as output, so cost per answer can diverge much further than cost per token.
What actually differs
- Kimi K3 has open weights under Kimi K3 License, so it can run on your own hardware with no per-token cost and no dependency on an API staying available. The other is API-only.
- Kimi K3 takes 1M tokens of context against 1M — 1.0× more room for long documents or a large codebase.
Scorecard
Which is better at what
Maths, coding, reasoning and the rest — one line each, averaged over the benchmarks both models actually report.
Kimi K3 has no published benchmark scores
Its maker has not released figures for any of the evaluations tracked here, so there is nothing to put in a score column. Rather than estimate, infer from a sibling model, or quote the base model's numbers as if they were its own, this page leaves those rows empty and compares what genuinely can be compared: parameters, context window, modalities, licence and cost.
The moment those figures are published they go in — send them with a link to the source.
| Category | Kimi K3 | Claude Opus 5 | Better at this |
|---|---|---|---|
| ReasoningOnly one model reports this | — | — | Not comparable |
| MathsNeither model reports this | — | — | Not comparable |
| CodingOnly one model reports this | — | — | Not comparable |
| KnowledgeNeither model reports this | — | — | Not comparable |
| MultimodalNeither model reports this | — | — | Not comparable |
| Instruction followingNeither model reports this | — | — | Not comparable |
| Human preferenceNeither model reports this | — | — | Not comparable |
| Categories won | No category has a test both models report | ||
Each category averages only the benchmarks every model here reports, so no one is credited for a test the other did not run. A category with no shared test is marked Not comparable rather than guessed at.
Side by side
Kimi K3 and Claude Opus 5, row by row
| Attribute | Kimi K3Moonshot AI | Claude Opus 5Anthropic |
|---|---|---|
| Specification | ||
| MakerWho built it | Moonshot AI | Anthropic |
| Released | 2026-07 | 2026-07 |
| ParametersTotal, and active per token for a mixture of experts | 2.8T total / 104B active | Not reported |
| Architecture | MoE | Not reported |
| Context windowHow much can go in at once | 1,048,576 tokens | 1,000,000 tokens |
| Max output | 1,048,576 tokens | 128,000 tokens |
| Input | Text, Image | Text, Image |
| ReasoningSpends extra tokens thinking before it answers | Yes | Yes |
| Tool calling | Yes | Yes |
| Knowledge cutoff | Not reported | 2026-05 |
| Licence | Kimi K3 License | Proprietary |
| Open weightsCan you download and run it yourself | Yes | No |
| Price | ||
| Input priceUSD per million tokens in | $3.00 | $5.00 |
| Output priceUSD per million tokens out | $15.00 | $25.00 |
| Cached input | $0.30 | $0.50 |
| Blended 3:1A 3-in-to-1-out million-token mix — a fairer single number than input price alone | $6.00Cheapest | $10.00 |
| Price note | Open weights — this is a representative hosting rate, not a first-party price. Running it yourself costs only hardware. | First-party API rate. |
| Published benchmarks | ||
| Terminal-Bench 2.1The 2.1 revision of the terminal agent benchmark. Scores on it are not comparable with Terminal-Bench 4.0 — the task set changed. | Not reported | |
| Frontier-Bench v0.1Novel problems built to resist memorisation, scored on whether the model gets anywhere at all. Absolute numbers are low by design. | Not reported | |
| Links | Hugging Face · Full page | Full page |
| Row verified | 2026-09 | 2026-09 |
Where these numbers come from
Every score on this page is a published figure, taken from the model's own card, system card, technical report or release post, or from a public leaderboard. CorX Labs did not run these evaluations. Most are self-reported by the lab that built the model, which means they were produced under that lab's own choice of prompt, scaffold and number of attempts — so treat them as a starting point for a shortlist, not as a settled ranking.
A score someone other than the model's maker measured is marked Independent and names its measurer. Those are the stronger numbers on this page — an outside harness has no reason to flatter anyone — and there are not many of them.
Where a figure has not been published, the cell reads Not reported rather than an estimate. Nothing here is inferred, interpolated or guessed. Each model records the month its row was last checked. Full method and caveats.
Questions
Kimi K3 or Claude Opus 5?
Is Kimi K3 better than Claude Opus 5?
These two models have no benchmark in common with published figures for both, so there is nothing to compare directly. The specification and price rows below are still like for like.
Which is cheaper, Kimi K3 or Claude Opus 5?
On a 3:1 input-to-output mix, Kimi K3 costs $6.00 per million tokens against $10.00 for Claude Opus 5 — about 1.7× less. Remember that a reasoning model bills its thinking as output, so cost per answer can diverge much further than cost per token.
What is the difference between Kimi K3 and Claude Opus 5?
Kimi K3 has open weights under Kimi K3 License, so it can run on your own hardware with no per-token cost and no dependency on an API staying available. The other is API-only. Kimi K3 takes 1M tokens of context against 1M — 1.0× more room for long documents or a large codebase.
Related
