Command Palette
Search for a command to run...

Moonshot’s Kimi K2.7 Code Places 2nd in 14-Problem ErdosBench Smoke Test, Ahead of GPT-5.5 xhigh

aiai-modelingai-research-evalsai-model-releasesai-open-models 31 posts · 21 accounts

In a rerun of the ErdosBench smoke test on 14 problems, Moonshot AI’s Kimi K2.7 Code ranked second behind Claude Fable‑5‑max and ahead of OpenAI’s GPT‑5.5 xhigh, according to ErdosBench’s published analysis. The benchmark said Kimi covered 13 of 14 problems, posted accepted solved or settled results on Problems 1, 3, 4, 5 and 7, and had no rejected solved claims.

The analysis described Kimi as the strongest new batch run and highlighted a Problem 3 result that upgraded the reference status to “superpolynomial but subexponential.” Claude Fable‑5‑max kept the top spot because it matched Kimi on the five accepted core results with full 14/14 coverage and broader accepted partial progress. The ranking adds an outside comparison point a day after Moonshot released and open-sourced Kimi-K2.7-Code, saying the model used 30% fewer reasoning tokens than K2.6.

From the sources (25 posts)

@kimi_moonshot

🌘 Kimi-K2.7-Code, our latest coding model, is now released and open-sourced! 🔷 Improved coding & agent performance over K2.6: +21.8% on Kimi Code Bench v2, +11.0% on Program Bench, and +31.5% on MLS Bench Lite. 🔷 Reasoning efficiency: Less

@kimi_moonshot

🎸 We're also launching the Kimi Code Beta Program today. Apply now if you'd like to try upcoming models and features before public release 👉

@stevibe

Kimi K2.7-Code!

@kimi_moonshot

🔗 Weights & code:

@teortaxestex

I want to see this compared with Composer 2.5 Like, really hard Cursor has a ton of proprietary data, a large head start, and threw a Colossus at RLing Kimi K2.5 checkpoint. What is the gap now?

@vanstriendaniel

RT @Kimi_Moonshot: 🔗 Weights & code:

@crystalsssup

less overthinking 👀

@teortaxestex

HUGE @htihle and others will get to test this soon I hope

@zephyr_z9

Great work from KIMI

@eliebakouch

RT @Kimi_Moonshot: 🌘 Kimi-K2.7-Code, our latest coding model, is now released and open-sourced! 🔷 Improved coding & agent performance over…

@code_star

It would be really cool if the top Chinese labs could properly benchmark and evaluate Fable/Mythos to show how it compares to their releases. I guess that isn’t possible though.

@thezachmueller

RT @Kimi_Moonshot: 🌘 Kimi-K2.7-Code, our latest coding model, is now released and open-sourced! 🔷 Improved coding & agent performance over…

@kimmonismus

Moonshot just released Kimi-K2.7 code, a huge upgrade to Kimi-K2.6! Big jump over K2.6: +21.8% on Kimi Code Bench v2 +11.0% on Program Bench +31.5% on MLS Bench Lite It also uses 30% fewer reasoning tokens, follows instructions better, and

@andrewcurran_

RT @Kimi_Moonshot: 🌘 Kimi-K2.7-Code, our latest coding model, is now released and open-sourced! 🔷 Improved coding & agent performance over…

@nrehiew_

I think K2.6 is the third best model on the planet - albeit one of the slowest. With the overthinking fixed and overall performance improvements, can’t wait to use this

@vllm_project

🎉 Congrats to @Kimi_Moonshot on Kimi K2.7-Code, a coding-focused agentic model built on K2.6. ✨ 1T-parameter Mixture-of-Experts, 32B active per token ✨ MLA attention with a 256K-token context window ✨ ~30% fewer thinking tokens than K2.6 f

@clementdelangue

RT @Kimi_Moonshot: 🔗 Weights & code:

@kimi_moonshot

RT @vllm_project: 🎉 Congrats to @Kimi_Moonshot on Kimi K2.7-Code, a coding-focused agentic model built on K2.6. ✨ 1T-parameter Mixture-of-…

@matthewberman

Nearly frontier open source coding model. Congrats to the Kimi team. I would love to see this tested against DeepSWE

@thezachmueller

RT @elliotarledge: I benchmarked Kimi K2.7-Code (1T MoE, coding-specialized, just dropped) on KernelBench-Hard, against its general predece…

@threepointone

RT @michellechen: k2.7 code live on workers ai

@thdxr

RT @opencode: Kimi 2.7 Code now available in Go text · image · optimized for coding similar pricing as 2.6

@teknium

Kimi 2.7 Coder now available in Hermes Agent! No need to update - your model picker will pick it up automatically.

@ollama

.@Kimi_Moonshot's kimi-k2.7-code is now available on Ollama's cloud! On Ollama's cloud, this model is hosted in the US on the latest NVIDIA B300 datacenter GPUs. Your data stays private and is never trained on. ❤️❤️❤️ Try the model: C

@testingcatalog

Kimi-K2.7-Code is now available on AI/ML API 👀 > Kimi K2.7 Code is the latest agentic coding model from Kimi AI that supports extended reasoning and tool use. > AI/ML API is a single gateway to Chat, Reasoning, Image, Video, Audio, Voice

Preview built on a synthetic news corpus (16 weeks, Apr–Jul 2026). Impact calls are model reads, not price data.

About Archive