NEWHow AI models draw the world map · Gecko Tests
Beta
Home/Models/Magnum v4 72B
anthracite-org logo

Magnum v4 72B

by anthracite-org · Released Oct 2024

Open Source
50.9
avg score
Rank #140
Compare
Better than 55% of all models
Context
33K tokens (~16 books)
Input $/1M
$3.00
Output $/1M
$5.00
Type
text
License
Open Source
Benchmarks
6 tested
Data as of
About

This is a series of models designed to replicate the prose quality of the Claude 3 models, specifically Sonnet(https://openrouter.ai/anthropic/claude-3.5-sonnet) and Opus(https://openrouter.ai/anthropic/claude-3-opus). The model is fine-tuned on top of [Qwen2.5 72B](https://openrouter.ai/qwen/qwen-

Tested on 6 benchmarks · BenchGecko score 50.9. Top scores: IFEval (56.3%), BBH (HuggingFace) (35.5%), MMLU-PRO (31.4%).

Looking for similar performance at lower cost?
Gemma 2 9B scores 51.1 (100% as good) at $0.03/1M input · 99% cheaper
Capabilities
reasoning
13.4
#157 globally
math
20.0
#223 globally
knowledge
20.9
#267 globally
language
56.3
#114 globally
general
35.5
#97 globally
Benchmark Scores
Compare All
Tested on 6 benchmarks · Ranked across 5 categories
Score Distribution (all 312 models)
0255075100
▲ You are here
reasoningCompare reasoning →
MUSR

HuggingFace MuSR (Multi-Step Reasoning). Tests multi-hop reasoning requiring chaining multiple facts together.

13.4·
mathCompare math →
MATH Level 5

HuggingFace evaluation of MATH Level 5 problems. Competition math requiring advanced reasoning and proof construction.

20.0·
knowledgeCompare knowledge →
MMLU-PRO

HuggingFace MMLU-Pro. Harder version of MMLU with 10 answer choices instead of 4 and more challenging questions.

31.4·
GPQA

HuggingFace evaluation of GPQA (Graduate-Level Google-Proof Q&A). PhD-level science questions that cannot be easily searched.

10.4·
Quick compare:Microsoft logovs Phi 3 Mini 4k InstructGoogle DeepMind logovs Gemini 2.0 ProGoogle DeepMind logovs Gemma 2 9B
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Similar Models
Microsoft logo
Phi 3 Mini 4k Instruct
Microsoft
50.8n/a
Google DeepMind logo
Gemini 2.0 Pro
Google DeepMind
50.7n/a
Google DeepMind logo
Gemma 2 9B
Google DeepMind
51.1$0.03/1M
Links
Info
anthracite-orgPricing explorerDevelopers · API
Research
Technical Report
Documentation
API DocsPlayground
Community
@anthracite-org
Source Code
GitHubHugging Face
BenchGecko API
magnum-v4-72b
Specifications
  • Typetext
  • Context33K tokens (~16 books)
  • ReleasedOct 2024
  • LicenseOpen Source
  • StatusActive
  • Cost / Message~$0.011
Available On
anthracite-org logoanthracite-org$3.00
Categories
reasoningmathknowledgelanguagegeneral
Learn More
transformeropen weightstokens
Share & Export
Tweet
Related Models
Microsoft logoPhi 3 Mini 4k Instruct
50.8
Google DeepMind logoGemini 2.0 Pro
50.7
Google DeepMind logoGemma 2 9B
51.1
Microsoft logoPhi 3.5 Mini Instruct
51.1
Magnum v4 72B is an open-source text AI model by anthracite-org, released in October 2024. It has an average benchmark score of 50.9. Context window: 33K tokens.

Related Models

Phi 3 Mini 4k Instruct · MicrosoftGemini 2.0 Pro · Google DeepMindGemma 2 9B · Google DeepMindPhi 3.5 Mini Instruct · MicrosoftMiniMax M2.7 · minimax

Benchmarks

IFEvalBBH (HuggingFace)MMLU-PROMATH Level 5MUSR

Related Pages

anthracite-org · Provideranthracite-org · EconomyAll ModelsCompare ModelsPricingDevelopers · API

Key facts · as of 2026-10-05

  • Magnum v4 72B by anthracite-org. BenchGecko score 50.9, rank 140 of 312 scored models (normalized average of public benchmark scores).
  • List price $3.00 input · $5.00 output per 1M tokens (as of 2026-10-05).
  • Sold by 1 provider (as of 2026-10-05): Mancer 2 (fp8) $2.50 in / $5.00 out.

How to cite · data as of 2026-10-05

Magnum v4 72B · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-05. https://benchgecko.ai/model/magnum-v4-72b

Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP