Granite-4.0-H-Small vs Sarvam 105B vs Apertus 8B
No independent capability benchmark covers every model here yet. The score uses price and context window.No independent capability benchmark covers every model here yet. The score uses context window.
IBM
Granite-4.0-H-Small
74/100- ECI—
- Price$0.064 / $0.265
- Context131K
Sarvam AI
Sarvam 105B
61/100- ECI—
- Price$0.158 / $0.632
- Context131K
Swiss AI
Apertus 8B
69/100- ECI—
- Price$0.10 / $0.20
- Context66K
IBM
Granite-4.0-H-Small
32/100- ECI—
- Price$0.064 / $0.265
- Context131K
Sarvam AI
Sarvam 105B
32/100- ECI—
- Price$0.158 / $0.632
- Context131K
Swiss AI
Apertus 8B
22/100- ECI—
- Price$0.10 / $0.20
- Context66K
Capability evidence is incomplete
There is not enough shared capability evidence to name an overall winner. The scores (Granite-4.0-H-Small 74/100, Apertus 8B 69/100, Sarvam 105B 61/100) compare the available specifications and price. The score weighs price 71%, context window 29%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.There is not enough shared capability evidence to name an overall winner. The scores (Granite-4.0-H-Small 32/100, Sarvam 105B 32/100, Apertus 8B 22/100) compare the available specifications only. The score weighs context window 100%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
- CapabilityNot enough dataNo independent benchmark covers every model here yet
- Lowest priceGranite-4.0-H-SmallGranite-4.0-H-Small $0.114 · Apertus 8B $0.125 · Sarvam 105B $0.276 per 1M tokens (3:1 blend)
- Longest contextGranite-4.0-H-Small and Sarvam 105BGranite-4.0-H-Small 131,072 · Sarvam 105B 131,072 · Apertus 8B 65,536 tokens
- Widest inputsSame inputsGranite-4.0-H-Small: Text · Sarvam 105B: Text · Apertus 8B: Text
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | Granite-4.0-H-Small | Sarvam 105B | Apertus 8B |
|---|---|---|---|---|
| Price | 71% | 90 | 73 | 89 |
| Context window | 29% | 32 | 32 | 22 |
| Specs & price only | 100% | 74/100 | 61/100 | 69/100 |
Left out because at least one model lacks the data: capability and inputs & features. The remaining weights were rescaled.
| Measure | Weight | Granite-4.0-H-Small | Sarvam 105B | Apertus 8B |
|---|---|---|---|---|
| Context window | 100% | 32 | 32 | 22 |
| Specs only | 100% | 32/100 | 32/100 | 22/100 |
Left out because at least one model lacks the data: capability and inputs & features. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Unavailable specifications are marked “Not published”. Publisher results include their evaluation settings and source.
| Specification | |||
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | Not published | Not published | Not published |
| ECI rank | Not published | Not published | Not published |
| ECI confidence interval | Not published | Not published | Not published |
| ECI evaluation date | Not published | Not published | Not published |
| Documented capabilities | Open-weight hybrid model for enterprise chat, coding, retrieval-augmented generation, and tool-calling workloads | Flagship Indian-language reasoning model for enterprise multilingual applications | Fully open 8B multilingual LLM supporting 1800+ languages with 65K context. Trained on compliant open data. Apache 2.0, EU AI Act compliant. |
| Model parameters | Not published | Not published | Not published |
| Output modalities | Text | Text | Text |
| Price per million tokens | |||
| Input | $0.064 (best) | $0.158 | $0.10 |
| Output | $0.265 | $0.632 | $0.20 (best) |
| Cached input | Not published | $0.053 | Not published |
| Blended (3:1) | $0.114 (best) | $0.276 | $0.125 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official watsonx.ai API | Official Sarvam AI API | Median of 1 providers |
| Limits | |||
| Context window | 131,072 tokens (best) | 131,072 tokens (best) | 65,536 tokens |
| Max input | Not published | Not published | Not published |
| Max output | 131,072 tokens (best) | 131,072 tokens (best) | 8,192 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | No | No |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | No | Yeslow · medium · high | No |
| Tool calling | Yes | Yes | Yes |
| Structured output | Yes | Not published | Not published |
| File attachments | No | No | No |
| Temperature control | Yes | Yes | Yes |
| Reasoning control | Not published | effort | Not published |
| Availability | |||
| Weights | Open | Open | OpenApache-2.0 |
| API model ID | ibm/granite-4-h-small | sarvam-105b | Not published |
| API providers | 1 | 3 (best) | 1 |
| Released | Oct 2, 2025 | Feb 18, 2026 | Sep 2, 2025 |
| Knowledge cutoff | Not published | Not published | Sep 2025 |
| Publisher documentation | Publisher documentation · Model weights | Publisher documentation · Model weights | Model weights |
| Source metadata updated | Oct 2, 2025 | Feb 18, 2026 | Sep 2, 2025 |
| Data sources | models.dev · Epoch AI | models.dev · Epoch AI | models.dev · Epoch AI |
| Catalog checked | Nov 22, 2026 | Nov 22, 2026 | Nov 22, 2026 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
Granite-4.0-H-Small$1.17
Sarvam 105B$2.84
Apertus 8B$1.40
Which should you choose?
Which is better: Granite-4.0-H-Small, Sarvam 105B or Apertus 8B?
There is not enough shared capability evidence to name an overall winner. The scores (Granite-4.0-H-Small 74/100, Apertus 8B 69/100, Sarvam 105B 61/100) compare the available specifications and price. The score weighs price 71%, context window 29%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.There is not enough shared capability evidence to name an overall winner. The scores (Granite-4.0-H-Small 32/100, Sarvam 105B 32/100, Apertus 8B 22/100) compare the available specifications only. The score weighs context window 100%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
Which is cheaper, Granite-4.0-H-Small, Sarvam 105B or Apertus 8B?
Granite-4.0-H-Small is cheaper at $0.064 input / $0.265 output per million tokens (official watsonx.ai API price). Apertus 8B costs $0.10 input / $0.20 output per million tokens (median across 1 API provider); Sarvam 105B costs $0.158 input / $0.632 output per million tokens (official Sarvam AI API price). At a typical mix of three input tokens to one output token, that is $0.114 per million tokens for Granite-4.0-H-Small versus $0.125 for Apertus 8B (1.1× as much) and $0.276 for Sarvam 105B (2.4× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. Granite-4.0-H-Small has not been scored yet, Sarvam 105B has not been scored yet and Apertus 8B has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for Granite-4.0-H-Small, Sarvam 105B and Apertus 8B yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. All three support tool calling for agent workflows.
Which has the bigger context window?
Granite-4.0-H-Small and Sarvam 105B have the largest context windows (131,072 and 131,072 tokens), against 65,536 for Apertus 8B. Maximum output per response: Granite-4.0-H-Small up to 131,072, Sarvam 105B up to 131,072, Apertus 8B up to 8,192 tokens.
Which can read images, PDFs, audio or video?
Granite-4.0-H-Small accepts text; Sarvam 105B accepts text; Apertus 8B accepts text. They handle the same number of input types.
Are any of these open source?
Yes, all three publish their weights (Apache-2.0), so you can self-host them.
Which is newer?
Sarvam 105B is the newest, released Feb 18, 2026. Granite-4.0-H-Small came out Oct 2, 2025; Apertus 8B came out Sep 2, 2025. Knowledge cutoff: Apertus 8B Sep 2025.
How do you decide the winner?
This rating uses price (71%, blended price per million tokens (3 input : 1 output), log scale); context window (29%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 22, 2026.This rating uses context window (100%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 22, 2026.