GeekRadar / The comparison desk

COMPARE.

AI models

allenai

OLMo-2-0325-32B-Instruct

AI model

Context window: 4,096 tokens
Total parameters: 32.23 B
Open weights: Yes

Capability & preference

Human preference · Arena

Overall text preferences from head-to-head votes. Not factual accuracy or a task-success rate.

OLMo-2-0325-32B-Instruct 1217.9 olmo-2-0325-32b-instruct · 95% interval 1207.1–1228.7 · 3,334 votes · 2026-10-15

Recorded results

Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.

Arena text preference / chinesehuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,191.47 rating

Confidence interval: 1,155.64 – 1,227.31

229 votes

Oct 15, 2026

Category: chinese
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / codinghuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,209.91 rating

Confidence interval: 1,182.68 – 1,237.15

427 votes

Oct 15, 2026

Category: coding
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,198.39 rating

Confidence interval: 1,173.73 – 1,223.06

567 votes

Oct 15, 2026

Category: creative_writing
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / englishhuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,255.37 rating

Confidence interval: 1,242.39 – 1,268.35

2025 votes

Oct 15, 2026

Category: english
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,106.76 rating

Confidence interval: 1,089.65 – 1,123.88

2276 votes

Oct 15, 2026

Category: exclude_ties
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,207.60 rating

Confidence interval: 1,187.92 – 1,227.28

789 votes

Oct 15, 2026

Category: hard_prompts
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,243.53 rating

Confidence interval: 1,219.92 – 1,267.13

538 votes

Oct 15, 2026

Category: hard_prompts_english
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,181.01 rating

Confidence interval: 1,148.97 – 1,213.06

336 votes

Oct 15, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,224.48 rating

Confidence interval: 1,201.38 – 1,247.58

606 votes

Oct 15, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,228.08 rating

Confidence interval: 1,187.80 – 1,268.36

202 votes

Oct 15, 2026

Category: industry_legal_and_government
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,239.47 rating

Confidence interval: 1,215.85 – 1,263.08

580 votes

Oct 15, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,195.37 rating

Confidence interval: 1,165.54 – 1,225.19

336 votes

Oct 15, 2026

Category: industry_mathematical
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,181.68 rating

Confidence interval: 1,137.53 – 1,225.82

204 votes

Oct 15, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,213.86 rating

Confidence interval: 1,192.44 – 1,235.28

740 votes

Oct 15, 2026

Category: industry_software_and_it_services
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,203.66 rating

Confidence interval: 1,184.30 – 1,223.02

872 votes

Oct 15, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,185.82 rating

Confidence interval: 1,168.60 – 1,203.04

1063 votes

Oct 15, 2026

Category: instruction_following
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / longer_queryhuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,193.90 rating

Confidence interval: 1,160.05 – 1,227.75

261 votes

Oct 15, 2026

Category: longer_query
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / mathhuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,207.55 rating

Confidence interval: 1,179.36 – 1,235.74

375 votes

Oct 15, 2026

Category: math
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,221.22 rating

Confidence interval: 1,192.10 – 1,250.34

363 votes

Oct 15, 2026

Category: multi_turn
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / non_englishhuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,160.24 rating

Confidence interval: 1,144.15 – 1,176.33

1309 votes

Oct 15, 2026

Category: non_english
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / overallhuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,217.92 rating

Confidence interval: 1,207.14 – 1,228.69

3334 votes

Oct 15, 2026

Category: overall
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Arena text preference / russianhuman preference · 2026-10-15 · rating

OLMo-2-0325-32B-Instruct

1,186.84 rating

Confidence interval: 1,154.15 – 1,219.53

304 votes

Oct 15, 2026

Category: russian
Evaluated configuration: olmo-2-0325-32b-instruct
Evaluation type: human preference

Run it locally

Inference · memory estimate

OLMo-2-0325-32B-Instruct

32.23 B total parameters · olmo2 · BF16

Published checkpoint tensor payload: 60.0 GiB. This includes stored copies and buffers; it is separate from the runtime estimates below.

Weight precision Total context Check a GPU Check a Local AI system

Estimated accelerator memory19.2 GiB

Weights: 16.5 GiB
KV cache: 1.0 GiB
Runtime allowance: 1.7 GiB

One user, full model in accelerator memory.

GPUs with memory headroom

Local AI systems with memory headroom

AMD Ryzen AI Halo (Ryzen AI Max+ 395, 128 GB) Planning budget: 96.0 GiB
ASUS Ascent GX10 (128 GB) Planning budget: 96.0 GiB
Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB) Planning budget: 96.0 GiB
Framework Desktop (Ryzen AI Max+ 395, 128 GB) Planning budget: 96.0 GiB
NVIDIA DGX Spark (128 GB) Planning budget: 96.0 GiB

Shared-memory systems use a planning budget of 75% of installed memory, leaving 25% for the OS and other allocations. This is an assumption, not a measured GPU allocation. Firmware and driver limits may be lower and may require configuration. Memory matches do not verify a runtime, quantized model artifact or inference speed.

Includes FP16 KV cache, prompt plus generated tokens, quantization metadata and a runtime allowance. GPU matches reserve 10% of VRAM. Memory fit is not a speed or software-compatibility guarantee; quantized weights and runtime support must be available.

Capabilities

Context window: 4,096 tokens
Open weights: Yes
Model license: apache-2.0
Total parameters: 32.23 B
Checkpoint tensor payload: 60.04 GiB

Input modalities

text

Output modalities

text

Processing & platform

Architecture: olmo2