Sarvam AI
sarvam-105b
apache_2_0
Benchmark Average (0–100)
56.2%
Average of non-null benchmark scores across all evaluated tasks.
BENCHMARK SCORES
GPQA78.7%
Graduate-Level Google-Proof Q&A. PhD-level multiple-choice questions in chemistry, biology, and physics. Scored by accuracy.
AIME 202596.7%
AIME 2025 is a mathematics competition benchmark testing advanced problem-solving.
SWE-bench Verified45.0%
SWE-bench Verified evaluates models on real-world software engineering tasks from GitHub issues.
BrowseComp49.5%
BrowseComp tests web browsing comprehension, measuring ability to find and synthesize information from websites.
HLE11.2%
Humanity's Last Exam (HLE) is a multi-modal benchmark testing frontier knowledge across mathematics, humanities, and natural sciences with 2,500 expert-level questions.
MODEL INFO
Organization
Sarvam AI
Context Window
—
Release Date
2026-03-06
Knowledge Cutoff
—
Parameters
105000000000.0B
License
apache_2_0
METADATA
Announcement Date
2026-03-06
Organization Country
IN
Canonical ID
—