Compare AI models
| Specifications | ||
ParametersA rough measure of how big the model is. More parameters usually means more capable and more expensive to run, though it is a poor guide on its own — a smaller, newer model often beats a larger, older one. | — | 1T |
Context windowHow much text the model can hold in mind at once — your question, any documents you attach, the conversation so far, and its own reply. Go past it and the earliest part falls out of view. | 1M | 256k |
| API pricingUSD per 1M tokens · lower wins · base tier | ||
Input priceWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it. | $0.10 | $0.60 |
Output priceWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one. | $0.50 | $3.00 |
Cached input priceA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again. | $0.01 | $0.10 |
Cheapest inputLowest input rate across third-party providers, excluding the lab itself. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | — | $0.45SiliconFlow |
Cheapest outputLowest output rate across third-party providers, excluding the lab itself. May come from a different provider than the cheapest input. | — | $2.25SiliconFlow |
| Benchmarks | ||
Humanity's Last Exam · with toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects. “With tools” means the AI is allowed to search the web or run code while answering. Higher is better. | 57.4% | 30.1% |
BullshitBench v2Nonsense detection — Given a confidently-worded but nonsensical prompt, does the AI spot that it makes no sense and push back — instead of playing along and inventing an answer? The score is how often it clearly called out the nonsense. Higher is better. | — | 52% |
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | — | 76.8% |
FrontierCode v1.1 (Main) · main splitAgentic coding — A set of very hard, frontier-difficulty coding tasks an AI agent has to complete end to end. The score is the share of tasks in the main split it solves. Higher is better. | 46.4% | — |
Next.js EvalsNext.js coding — Vercel's open eval of how well AI coding agents build and migrate real Next.js apps — measured as the share of tasks the agent completes successfully. Higher is better. | — | 16% |
LiveCodeBenchCompetitive coding — Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better. | — | 85% |
Terminal-Bench 4.0Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Version 4.0 recalibrated how much time, CPU and memory each task gets, removed eight tasks and fixed nineteen, so fewer runs fail for reasons that have nothing to do with the model. Scores are not comparable with earlier versions. Higher is better. | 39.2% | — |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | — | 60.6% |
Humanity's Last Exam · no toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects, written so you can't just look up the answer. “No tools” means the AI answers on its own. Higher is better. | 45.9% | — |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | — | 87.6% |
OSWorld 2.1 (offline)Agentic computer use — Can the AI actually operate a computer — clicking, typing, and using real apps — to finish tasks on its own? Version 2.1, in its offline setting. Scores are not comparable with version 2.0. Higher is better. | 72.4% | — |
GDPval-AA v2.1Knowledge work — economically valuable knowledge work (v2.1, Crowd-BT Elo fit) | 1620 | — |
AA-Briefcase v1.1Knowledge work — Artificial Analysis agentic office-work eval (Elo, v1.1 rating fit) | 1578 | — |
Chartography · no toolsChart tasks — The same chart-centred test with no tools: the AI has to read each chart unaided. Scores run far lower than the with-tools version, so read the two as separate tests. Higher is better. | 46.4% | — |
| Overview | ||
| Company | Anthropic | Moonshot AI |
| Release date | Oct 7 2026 | Jan 27 2026 |
| Access | Closed | Open Weight |
| Model details | View model | View model |
Other comparisons
Claude Haiku 5.5vsGPT-6.1 SolKimi K2.5vsGPT-6.1 SolClaude Haiku 5.5vsGemini 4 ArgonKimi K2.5vsGemini 4 ArgonClaude Haiku 5.5vsMuse Spark 1.3Kimi K2.5vsMuse Spark 1.3Claude Haiku 5.5vsGrok 4.7Kimi K2.5vsGrok 4.7Claude Haiku 5.5vsDeepSeek-V4.1-FlashKimi K2.5vsDeepSeek-V4.1-FlashClaude Haiku 5.5vsMistral Large 4Kimi K2.5vsMistral Large 4Frequently asked questions
Claude Haiku 5.5 leads Kimi K2.5 on 1 of the 1 benchmark they both report (Humanity's Last Exam). Claude Haiku 5.5 is cheaper on both input and output: $0.10 vs $0.60 per million input tokens, and $0.50 vs $3.00 per million output tokens. Figures are base-tier rates. Kimi K2.5 shipped 253 days before Claude Haiku 5.5, so benchmark comparisons should account for the intervening progress.
Context windows are 1M (Claude Haiku 5.5) vs 256k (Kimi K2.5). Claude Haiku 5.5 is closed, while Kimi K2.5 is open weight.
On Humanity's Last Exam · with tools, Claude Haiku 5.5 leads at 57.4% vs Kimi K2.5 at 30.1%.