How does Kimi K2.5 rank among AI models?
Kimi K2.5 is a superseded version and is not on the current board as of 22 September 2026. Its archive position, where one exists, is shown on the page; the newer version carries the current rank.
Moonshot AI·moonshotai/kimi-k2.5
Kimi K2.5 is ModelCap archive rank #62 (Measured), from the ModelCap Index as of 22 Sept 2026, 02:49 UTC.
Kimi K2.5 listed OpenRouter input price is $0.45 per 1M tokens and listed output price is $2.25 per 1M tokens, from OpenRouter listed prices as of 22 Sept 2026, 02:49 UTC.
Kimi K2.5 has a published context limit of 262K tokens in the ModelCap catalogue as of 22 Sept 2026, 02:49 UTC.
Kimi K2.5 has 5 published OpenRouter serving endpoints, from OpenRouter as of 22 Sept 2026, 02:49 UTC.
Kimi K2.5 weight access is restricted license, from ModelCap classification of published repository metadata as of 22 Sept 2026, 02:49 UTC.
Snapshot facts · download the public dataset · methodology
ModelCap Index · All-versions archive
71.4#62
Measured · 67% support · score interval 66.8–76.0
Evidence as of 21 Sept 2026, 17:21 UTC
Superseded by Kimi K3. It no longer holds a current-board position; its all-versions archive rank is #62.
Input
$0.45
per 1M tokens
Output
$2.25
per 1M tokens
Context
262K
tokens
Providers
5
endpoints
Gateway spend
—
latest · Vercel share
Providers
Uptime measured over the last 30 minutesHugging Face Inference Providers
Exact repository identityProvider economics and telemetry published by the Hugging Face Router. These are operational signals, not benchmark results.
Reasoning efforts
Best supported result can inform ModelCapPublished Max, X-High, High and other configurations for this model. These are sourced scores, not head-to-head essays.
Related community configurations
Human preference ratingReported evaluation leads (88)
Not independent measurementStructured rows reported in this exact model repository at the pinned revision below. They are the publisher talking about itself: never measured evidence, never a score input on their own. They remain quarantined until dataset revision, harness, configuration and identity pass ModelCap’s independent admission review.
Showing 20 of 88 rows.
Other versions (4)
ModelCap Index · All-versions archive
Methodology71.4
Measured · 4 public benchmark observations across 4 boards
All-versions archive
#62
- Public rank basis
- Measured (measured)
- Index support
- 66.9% · strong
- Index interval
- 66.8–76.0
- Rank posterior
- Not available
- Top 5 / Top 10 probability
- Not available / Not available
- Posterior as of
- Snapshot timestamp unavailable
- Evidence as of
- 21 Sept 2026, 17:21 UTC
- Identity binding
- Exact catalogue product · moonshotai/moonshotai/kimi-k2.5 · aggregates disclosed benchmark configurations · endpoint configuration not claimed · artifact metadata moonshotai/Kimi-K2.5@4d01dfe0332d63057c186e0b262165819efb6611 (not the evaluation revision)
- Index observations used
- 4
- Qualified Index evidence point
- 71.4
Bayesian Evidence Score · measured evidence input only
BES normalizes admitted public benchmark evidence for measured Index rows. Its legacy score and observables prior do not define the public language rank.
- Observed capability
- 71.4
- Evidence status
- confirmed · 90% mass
- General preference
- 73.6
- Coding
- 71.9
- Agents & tools
- 27.1
- Reasoning
- 35.9
- Evidence breadth
- 100%
- Evidence coverage
- 67%
Market and catalogue signals · secondary, never the language Index rank
- Usage (OpenRouter popularity)
- 21.3
- Liquidity (providers × uptime)
- 99.3
- Open reach (HF downloads)
- —
- Surface (context / tools / modalities)
- 90.3
- Freshness
- 40.4
- · OpenRouter popularity #94
- · 5 live providers
- · Catalogue: context 262144, reasoning, tools, modalities Text/Image
- · First seen 2026-01-27
These adoption and deployment observations remain context only. They do not change this model's ModelCap Index score or rank.
BES evidence selected · measured input only
- Arena Overall · Thinking73.6
- Arena Coding · Thinking71.9
- WildClawBench OpenClaw27.1
- ARC-AGI-2 Verified35.9
The bounded 0–100 capability score blends each source’s competitive placement with its published achievement, then combines capability families. Evidence breadth adds a modest uncertainty adjustment; price and popularity are not part of benchmark-led rank.
Deployment readiness
Metadata index86.3
strong · 89% metadata coverage
- Artifact reproducibility
- 85.0
- Access & legal clarity
- 67.5
- Deployability
- 97.6
- Evaluation provenance
- 89.9
Missing: artifactReproducibility.base-lineage, accessLegalClarity.resolvable-license-terms, evaluationProvenance.evaluation-dataset-revisions.
A metadata completeness and deployability index—not a safety certification, quality grade, or production-readiness claim.
Market activity
Venue coverage- Market Gravity
- 46.1
- OpenRouter weekly popularity
- #94 of 322
Market Gravity uses OpenRouter’s full-catalogue popularity order. Sparse Vercel top-ten observations appear only when published and do not affect the score.
Arena preference
- Status
- Ranked
- Source
- LMArena
- Official rank
- #69 of 402
- Rating
- 1450.4
- 95% confidence interval
- 1447–1454
- Votes
- 71K
- Source published at
- 13 Sept 2026, 00:00 UTC
- Category
- overall
- Identity match confidence
- 100%
Market Gravity breakdown
- Usage55%
- 21.3
- Liquidity25%
- 99.3
- Open reach15%
- 50.0
- Freshness5%
- 40.4
OpenRouter weekly popularity across the full model catalogue.
Independent providers versus the model's open or closed cohort, weighted by uptime.
Hugging Face 30-day downloads for open models; neutral for closed models.
Time since first public availability, on a six-month half-life.
Specification
- Context window
- 262K tokens
- Max output
- 236K tokens
- Inputs
- Text, Image
- Outputs
- Text
- Tokenizer
- Other
- Cached input tokens
- $0.07 / 1M
- First seen on OpenRouter
- 27 Jan 2026
Capabilities
- Supported: Reasoning
- Supported: Tool use
- Supported: Structured output
- Supported: Response format
- Not supported: Moderated
Weights & access
Restricted license. Downloadable weights use a model-specific or use-restricted license. Review the linked license before use, redistribution, or commercial deployment.
- Access
- Restricted license
- Licence
- modified-mit
- Pinned revision
- 4d01dfe0332d
- Parameters
- 1026.9B
- Architecture
- KimiK25ForConditionalGeneration
- Model type
- kimi_k25
- Hugging Face downloads (30d)
- 313K
- Downloads (all time)
- 15.8M
- Likes
- 3K
- Repository
- moonshotai/Kimi-K2.5
- Repository updated
- 30 Apr 2026, 03:56 UTC
Benchmark leaderboards featuring Kimi K2.5
All leaderboardsKimi K2.5: common questions
How much does Kimi K2.5 cost per 1M tokens?
Kimi K2.5 is listed at $0.45 per 1M input tokens and $2.25 per 1M output tokens as of 22 September 2026 across 5 providers. Prices come from the OpenRouter catalogue and re-observe on every refresh. For 1,000 requests using 2,000 input and 500 output tokens each, the listed rates imply $2.02 total (2M input + 0.5M output). This excludes caching, batch discounts, prompt-length tiers, tool charges and retries; verify the chosen endpoint's rate and limits before budgeting.
What is the context window of Kimi K2.5?
Kimi K2.5 has a published context window of 262K tokens (262,144) and a maximum output of 236K tokens. Individual providers can serve less than the published maximum; the providers table lists each endpoint's own limit.
Is Kimi K2.5 open-weight?
Restricted license (modified-mit). Downloadable weights use a model-specific or use-restricted license. Review the linked license before use, redistribution, or commercial deployment. The repository is moonshotai/Kimi-K2.5 on Hugging Face.
Which API providers serve Kimi K2.5?
5 providers serve Kimi K2.5 through OpenRouter as of 22 September 2026: SiliconFlow, AtlasCloud, Venice, Novita and Amazon Bedrock. Each provider's price, context limit, quantization and measured uptime are in the providers table above.
Which benchmarks has Kimi K2.5 been evaluated on?
Kimi K2.5 has published results on 6 public boards tracked by ModelCap as of 22 September 2026: Arena coding 1502 (#64 of 397), ARC-AGI-2 11.8% (#109 of 214), SWE-bench 70.8% (#12 of 47), AA τ²-Bench Telecom 95.9% (#16 of 436) and AA Humanity's Last Exam 30.7% (#111 of 615) and more. Each board page ranks every tracked model on that benchmark; the ModelCap Index combines them with published uncertainty.
When was Kimi K2.5 released?
Kimi K2.5 first appeared in the catalogue on 27 Jan 2026. It has since been superseded by Kimi K3. Rank and price movements since then are recorded on the site's changes feed.