Soofi S 30B-A3B

Soofi ProjectOpen weightsNon-reasoningReleased Jul 10, 2026

Compare with Claude Opus 5.5 Source scores ↗

Overall score

—

Not enough evidence to rank yet

Context window

1,000,000 tokens

Price per 1M tokens

— in · — out

No price listed

Blended price

—

3:1 input:output tokens

Output speed

—

Not measured yet

Category scores

Scores per capability area with the model's rank among ranked models in that category.

1Knowledge45.4

Higher is better. #N is the model's rank in that category, where it is eligible to be ranked.

Benchmark results (7)

As reported in system cards, papers and public evaluations. Position compares this model with every other model reporting the same benchmark.

Knowledge

GPQA	43.4	#106 of 117
MMLU-Pro	51.4	#58 of 67
AGIEval	66.9	#1 of 1

Reasoning

BBH	78.8	#2 of 7
DROP	66.5	#5 of 5

Coding

HumanEval

73.8

#13 of 18

Math

GSM8K

86.1

#5 of 6

Pricing by source

USD per 1M tokens.

Source	Input	Output	Cached in
List price	—	—	—

The Soofi Project publishes the Soofi S 30B-A3B base weights under Apache 2.0 for local and self-hosted use. The repositories remain access-gated during the beta phase, and no first-party hosted API token price is published, so BenchLM represents the checkpoint as self-host/free-per-token before infrastructure costs.

Speed

Measured by us through OpenRouter (median of trials).

Not available on OpenRouter, so we cannot measure it yet.