kimi-k2p5: benchmark scores, price and ranking

kimi-k2p5 from Moonshot AI ranks #289 of 743 LLMs on the LLMs Tiger aggregate benchmark score (49.2), based on 1 benchmark results captured up to 2026-10-13.

Organisation: Moonshot AI
Type: Open-weight / local model
Open weights: Yes
Cheapest API price: $0.600 in / $3.00 out per million tokens (via fireworks_ai, observed 2026-10-13T21:47Z)
Aggregate rank: #289 of 743 (score 49.2)

Across the 1 benchmark where at least ten models are scored, kimi-k2p5 is in the top 10% on 0 and in the top half on 0.

Its cheapest listed input price, $0.600 per million tokens, is lower than that of 45% of the 427 models with a listed price.

kimi-k2p5 benchmark results

kimi-k2p5 benchmark scores, ranked against all models measured on each benchmark
Benchmark	Score	Rank	Source	Captured
MCP Atlas	64.4%	#10 of 14	Scale Labs MCP Atlas leaderboard	2026-10-13

About kimi-k2p5

kimi-k2p5 is an open-weight model that can run on your own hardware from Moonshot AI. The cheapest API listing we track charges $0.600 per million input tokens and $3.00 per million output tokens, through fireworks_ai. We hold 1 benchmark result for it across 1 area, and none is in the top tenth of its board.

Where kimi-k2p5 stands

On our aggregate score, kimi-k2p5 is #289 of 743 ranked models, around the middle of the table. Its score is 49.2; the scale is centred on 50, so it sits 0.8 points below the typical model. That beats 61% of the ranked field. Among the 3 ranked Moonshot AI models it is #2. The next one up is Kimi K2 at 51.8. The next one down is kimi-k2.6-thinking at 48.2.

kimi-k2p5, benchmark by benchmark

kimi-k2p5 lands below the middle on MCP Atlas: 64.4%, #10 of 14, with 64% of the field scoring higher. Muse Spark leads at 82.2%; the closest result above is GPT-5.2 (xhigh) at 67.6%. It covers tool use across 1,000 MCP tasks. The board median is 70.4%, so kimi-k2p5 is 6.0 points below it. Source: Scale Labs MCP Atlas leaderboard, captured 2026-10-13.

Price and value

The median input price across the 427 priced models is $0.500 per million tokens, so kimi-k2p5 costs more than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), DeepSeek V4 Flash 0731 (max) ($0.016, 53.1). Similar scores cost more on Kimi K2.6 ($0.650) and kimi-k2.6-thinking ($0.650).

Reading these numbers

The 1 score on this page comes from 1 source: Scale Labs MCP Atlas leaderboard (1). 1 of them come from an official leaderboard and 0 from datasets or third-party runs. With 1 results the aggregate rank can move a lot when one new score arrives.

Models ranked nearby

#286 computer-use-preview-2025-03-11 (49.2)
#287 NVIDIA-Nemotron-3-Nano-30B-A3B-Base-BF16 (49.2)
#288 Seed-OSS-36B-Base (49.2)
#290 gpt-5-chat (49.2)
#291 Claude Code + Qwen3.8-Max-0902 (49.1)
#292 Claude Opus 4.1 (unknown thinking) (49.1)