Model reference · open weights

K2

Available as managed deployment LLMs IFM Text gen 1 variants 599 dl/mo

K2 is an open-weight language model from IFM. AxForge deploys and operates it for you on dedicated EU-owned hardware — with the licence handled where one is required.

Available as managed deployment — configured and operated for you on dedicated EU hardware, quoted per deployment.

What it is

Released byIFM
TypeLanguage models
TaskText gen
Parameters (lead)65.3B
Context8k tokens
Runs withtransformers
Released2024-04-17
Popularity599 downloads / month
LicenceOpen weights

About

What K2 is

LLM360 demystifies the training recipe used for Llama 2 70B with K2. K2 is fully transparent, meaning we’ve open-sourced all artifacts, including code, data, model checkpoints, intermediate results, and more.

Read the full model card

About K2:

  • 65 billion parameter LLM
  • Tokens: 1.4T
  • Languages: English
  • Models Released: base, chat model
  • Trained in 2 stages
  • License: Apache 2.0

K2 was developed as a collaboration between MBZUAI, Petuum, and LLM360.

LLM360 Model Performance and Evaluation Collection

The LLM360 Performance and Evaluation Collection is a robust evaluations set consisting of general and domain specific evaluations to assess model knowledge and function.

Evaluations include standard best practice benchmarks, medical, math, and coding knowledge. More about the evaluations can be found here.

Detailed analysis can be found on the K2 Weights and Biases project here

Open LLM Leaderboard

EvaluationScoreRaw Score
IFEval22.5223
BBH28.2250
Math Lvl 52.042
GPQA3.5828
MUSR8.5540
MMLU-PRO22.2730
Average14.5335.17

K2 Gallery

The K2 gallery allows one to browse the output of various prompts on intermediate K2 checkpoints, which provides an intuitive understanding on how the model develops and improves over time. This is inspired by The Bloom Book.

View K2 gallery here

Datasets and Mix

The following data mix was used to train K2 and achieve results in line with Llama 2 70B.

The full data sequence can be found here

DatasetStarting TokensMultiplierTotal Tokens% of Total
dm-math4.33B3x13B1%
pubmed-abstracts4.77B3x14.3B1.1%
uspto4.77B3x14.3B1.1%
pubmed-central26B1x26B2%
redpajama.arxiv27.3B1x27.3B2.1%
starcoder.spm67.6B0.5x33.8B2.6%
starcoder.fim67.6B0.5x33.8B2.6%
redpajama.stackexchange61.1B1x61.1B4.7%
starcoder132.6B0.5x66.3B5.1%
pile-of-law76.7B1x76.7B5.9%
redpajama.book80.6B1x80.6B6.2%
s2orc107.9B1x107.9B8.3%
redpajama.wikipedia22.1B6x132.6B10.2%
refinedweb612.3B1x612.3B47.1%
Totals--1.3T100%

LLM360 Reasearch Suite

Stage 2 - Last 10 Checkpoints

Checkpoints
Checkpoint 380Checkpoint 375
Checkpoint 379Checkpoint 374
Checkpoint 378Checkpoint 373
Checkpoint 377Checkpoint 372
Checkpoint 376Checkpoint 371

Stage 1 - Last 10 Checkpoints

Checkpoints
Checkpoint 360Checkpoint 355
Checkpoint 359Checkpoint 354
Checkpoint 358Checkpoint 353
Checkpoint 357Checkpoint 352
Checkpoint 356Checkpoint 351

[to find all branches: git branch -a]

LLM360 Pretraining Suite

We provide step-by-step reproducation tutorials for tech enthusiasts, AI practitioners and academic or industry researchers who want to learn pretraining techniques here.

LLM360 Developer Suite

We provide step-by-step finetuning tutorials for tech enthusiasts, AI practitioners and academic or industry researchers [here](https://www.llm360.ai/develo

From the published model card. Full card on the HuggingFace links in the sidebar.

Using it via the API

Call it like any OpenAI endpoint

Once AxForge deploys k2 for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (k2 below is illustrative; you get the exact model name on deployment.)

$ curl -sS https://api.axforge.ai/v1/chat/completions \
  -H "Authorization: Bearer $AXFORGE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"k2","messages":[{"role":"user","content":"Hello"}]}'

Create an account — your API key is available in the console. 3M free tokens every 30 days with every new account.

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms