Model

OLMo 2 7B

OPEN

Allen Institute for AI (Ai2) · OLMo 2 family · released Nov 26, 2024

Fully open 7B trained on up to 5T tokens; outperforms Llama 3.1 8B on English academic benchmarks despite lower training FLOPs.

ReasoningCodingVisionFunction callingTool useAgentic

446.1

Elo · rank #319

Parameters

7.3B

Active params

7.3B (dense)

Context

4K tokens

Architecture

Dense transformer (decoder-only)

License

Apache 2.0

Languages

API price (in/out)

No hosted API

Modalities

text

Benchmark results

Bar shows position within the tracked field; marker = field best

AGIEvalReasoning50.4%#16

best: OLMo 3-Think 32B · 88.2%

AlpacaEvalHuman preference29.1%#32

best: Tülu 2+DPO 70B · 95.1%

ARC-ChallengeReasoning79.8%#39

best: Llama 3.1 405B · 96.9%

DROPReasoning60.8#50

best: Hunyuan-T1 · 93.1

GSM8KMath67.5%#103

best: Llama 3.1 405B · 96.8%

HellaSwagReasoning83.8%#54

best: Claude 3 Opus · 95.4%

MATH-500Math32.5%#159

best: GPT-5 · 99.4%

MMLU-ProKnowledge31.0%#164

best: Claude Fable 5 · 91.5%

MMLUKnowledge63.7%#185

best: OpenAI o3 · 92.9%

TriviaQAKnowledge78.0%#15

best: Sarvam-1 (2B) · 90.6%

TruthfulQAKnowledge56.5%#33

best: Phi-3.5-MoE (16x3.8B, 6.6B active) · 77.5%

WinoGrandeReasoning77.2%#49

best: PaLM 2 · 90.9%

Run it locally

VRAM @ Q4

5 GB

VRAM @ FP16

15 GB

Fits on (Q4)

RTX 3060 12GBRTX 4070 Ti 16GBRTX 3090 24GBRTX 4090 24GBRTX 5090 32GBM4 Pro 48GBM3 Max 128GBM3 Ultra 512GBA100 80GBH100 80GBH200 141GBB200 192GB

Throughput data unavailable.

Quantizations

GGUF Q4

OLMo 2 family

Elo progression across releases

OLMo 2 13BNov 2024632.7 OLMo 2 7BNov 2024446.1 OLMo 2 32BMar 2025823.2

Resources

HFallenai/OLMo-2-1124-7B GHallenai/OLMo-core DOCSallenai.org/blog/olmo2 APIAvailable via Allen Institute for AI (Ai2) API

API price weights · each benchmark row carries its own source badge (see methodology)