Moonshot AI

• Moonshot AI, Kimi models, open-weight agentic performance at low cost.
• - Kimi K2 is the open-weight model most cited as competitive with closed frontier on agentic and tool-use benchmarks
• Released under a permissive modified-MIT licence
• First-party platform is inexpensive and OpenAI-compatible
• Weights also hosted by most routers here, the practical route if you cannot send data to a mainland-China endpoint.
• Founded 2023.
• Open source under Modified MIT.

• We track it in 1 comparison, LLM APIs.
• Every claim below is a cell in a table you can open and check rather than an impression.
• Across those rosters it sits against 17 other tools, and what follows is where it visibly separates from them.

Where it gives ground.

• Zero retention, No. Amazon Bedrock records Yes.
• Audio in/out, No. OpenAI records Partial.

• None of these disqualify it on their own.
• They are the fields to check against your own requirements before you commit, because they are the ones where a competitor genuinely does better.

• Provenance. 22 of 27 tracked fields carry a value for Moonshot AI, and 10 of those cite a document you can open.
• Last verified 2026-07-30.
• Every figure keeps its own provenance, measured by us, claimed by the vendor, inferred, or community-reported, and we would rather print a dash than a guess.

• Its nearest neighbour in our data is Qwen.
• Moonshot AI is ahead on Image input (Yes against Partial) and Open weights (Yes against Partial).
• Qwen takes MCP support (Partial against No) and AU region (Partial against No).
• That pattern repeats across the rest of the roster, see Moonshot AI alternatives for the other rivals, each compared the same way.

Founded

2023

Licence

Modified MIT (open source)

Fields we track

22 of 27

Last verified

2026-07-30

Kimi K2 is the strongest case that open weights caught up on agentic work; reach it through Groq, Fireworks or Together.

Zero retention

○No10th of 12

Inferredverified 2026-01-15

$/M output

$15/M tok7th of 9

Vendor-claimedverified 2026-07-30source

$/M input

$3/M tok7th of 9

Vendor-claimedverified 2026-07-30source

Tool use

●Yes1st of 18

Vendor-claimedverified 2026-07-30source

$/M cache read

$0.3/M tok5th of 8

Vendor-claimedverified 2026-07-30source

Pinnable versions

●Yes1st of 18

Inferredverified 2026-01-15

Context window

1,048,576tokens1st of 12

Vendor-claimedverified 2026-07-30source

OpenAI-compat API

●Yes1st of 18

Vendor-claimedverified 2026-07-30source

Moonshot AI is the right call in these situations, each one drawn from a field we actually record.

• Agent loops on an open-weight budget.
• Self-hosting a tool-use-capable model under a permissive licence.
• Replacing a frontier model on the cheaper half of a mixed workload.

Reach for something else when any of the following is a requirement rather than a nice-to-have.

• Zero retention. Moonshot AI records No on the LLM APIs table. Amazon Bedrock records Yes on the same field. If that is a hard requirement rather than a preference, start elsewhere.
• Audio in/out. Moonshot AI records No on the LLM APIs table. OpenAI records Partial on the same field. If that is a hard requirement rather than a preference, start elsewhere.

• We publish this block because a comparison that only lists what a tool is good at is marketing.
• Every figure above sits on the same page as its source, and the field definitions are on the category tables if you want to check how we measured them.

What we don't know.

• 5 of the 27 fields we track for Moonshot AI are still blank: Max output, TTFT p50, Output tok/s, Batch discount and Notice period.
• Those render as dashes rather than as zeroes or assumptions, because an empty cell and a bad cell are not the same thing and only one of them is honest.
• If you know any of these figures and can point at a document, tell us.