Sarvam 105B (high) Intelligence, Performance & Price Analysis
Model summary
Intelligence
Speed
Input Price
USD per 1M tokens
Output Price
Verbosity
Sarvam 105B (high) is above average in intelligence and well priced when comparing to other open weight models of similar size. It's also faster than average, however very verbose. The model supports text input, outputs text, and has a 128k tokens context window.
Sarvam 105B (high) scores 18 on the Artificial Analysis Intelligence Index, placing it above average among comparable models (averaging 15). When evaluating the Intelligence Index, it generated 71M tokens, which is very verbose in comparison to the average of 16M.
Pricing for Sarvam 105B (high) is $0.00 per 1M input tokens (competitively priced, average: $0.15) and $0.00 per 1M output tokens (competitively priced, average: $0.57). In total, it cost $0.00 to evaluate Sarvam 105B (high) on the Intelligence Index.
At 133 tokens per second, Sarvam 105B (high) is faster than average (73).
| Reasoning | Yes This page shows the reasoning version of this model. A non-reasoning variant may also exist. |
|---|---|
| Input modality | Supports: text |
| Output modality | Supports: text |
| Context window | 128k ~192 A4 pages of size 12 Arial font |
| Total parameters | 106B |
| Active parameters | 10.3B Number of parameters active per token during inference |
| License | Apache 2.0 |
| Model weights | Hugging Face |
Metrics are compared against models of the same class:
- Non-reasoning models → compared only with other non-reasoning models
- Reasoning models → compared across both reasoning and non-reasoning
- Open weights models → compared only with other open weights models of the same size class:
- Tiny: ≤4B parameters
- Small: 4B–40B parameters
- Medium: 40B–150B parameters
- Large: >150B parameters
- Proprietary models → compared across proprietary and open weights models of the same price range, using a blended 3:1 input/output price ratio:
- <$0.15 per 1M tokens
- $0.15–$1 per 1M tokens
- >$1 per 1M tokens
Highlights
Intelligence
Artificial Analysis Intelligence Index
Artificial Analysis Intelligence Index by Open Weights / Proprietary
Intelligence Evaluations
Agentic real-world work tasks, (ELO-500)/2000
Agentic coding & terminal use
Agentic tool use
Long context reasoning
Knowledge
1 - hallucination rate
Reasoning & knowledge
Scientific reasoning
Coding
Instruction following
Physics reasoning
Long-horizon agentic tasks
Visual reasoning
Openness
Artificial Analysis Openness Index: Results
Intelligence Index Comparisons
Intelligence vs. Price
Intelligence Index Token Use & Cost
Output Tokens Used to Run Artificial Analysis Intelligence Index
Cost to Run Artificial Analysis Intelligence Index
Context Window
Context Window
Pricing
Pricing now includes a “Cache Hit Price” alongside Input and Output pricing, with new blend ratios.
Pricing: Cache Hit, Input, and Output
Speed
Measured by Output Speed (tokens per second)
Output Speed
Output Speed vs. Price
Latency
Measured by Time (seconds) to First Token
Latency: Time To First Answer Token
End-to-End Response Time
Seconds to output 500 tokens, calculated based on time to first token, 'thinking' time for reasoning models, and output speed
End-to-End Response Time
Model Size (Open Weights Models Only)
Model Size: Total and Active Parameters
Frequently Asked Questions
Common questions about Sarvam 105B (high)
Sarvam 105B (high) was released on March 6, 2026.
Sarvam 105B (high) was created by Sarvam.
Sarvam 105B (high) scores 18 on the Artificial Analysis Intelligence Index, placing it above average among other open weight models of similar size (median: 15).
Sarvam 105B (high) generates output at 132.6 tokens per second (based on the median across providers serving the model), which is above average compared to other open weight models of similar size (median: 72.8 t/s).
Sarvam 105B (high) has a time to first token (TTFT) of 2.23s (based on the median across providers serving the model), which is somewhat higher than average compared to other open weight models of similar size (median: 1.89s).
When evaluated on the Intelligence Index, Sarvam 105B (high) generated 71M output tokens, which is somewhat higher than average compared to other open weight models of similar size (median: 16M).
Yes, Sarvam 105B (high) is a reasoning model. It uses extended thinking or chain-of-thought reasoning to work through complex problems before providing an answer.
Sarvam 105B (high) supports text input.
Sarvam 105B (high) supports text output.
No, Sarvam 105B (high) does not support image input. It can only process text.
No, Sarvam 105B (high) is not multimodal. It only supports text input.
Sarvam 105B (high) has a context window of 130k tokens. This determines how much text and conversation history the model can process in a single request.
Yes, Sarvam 105B (high) is open weights. The model weights are publicly available and can be downloaded for self-hosting.
Sarvam 105B (high) has 106 billion parameters (10.3 billion active).
Sarvam 105B (high) is a Mixture of Experts (MoE) model with 106 billion total parameters, but only 10.3 billion active parameters are used during inference.
Sarvam 105B (high) is released under the Apache 2.0 license. This license allows commercial use. View license
Sarvam 105B (high) achieves a score of 18 on the Artificial Analysis Intelligence Index. This composite benchmark evaluates models across reasoning, knowledge, mathematics, and coding.
Sarvam 105B (high) is an open weights model that can be self-hosted. View providers
Sarvam 105B (high) is an open weights model that can be downloaded and self-hosted. Compare providers