# Inference hosting: which vendors AI models recommend (September 2026)

Source: Memetik Index, https://www.memetik.ai/index/inference-hosting
Field: AI infra. Edition: 2026-09. Run: 2026-09-02.
Method: 5 fixed buyer prompts sent once each to 10 current models (GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, Claude Opus 5, Claude Sonnet 5, Claude Fable 5, Gemini 3.6 Flash, Gemini 3.5 Flash, Sonar Pro, Sonar Reasoning Pro) via their APIs with web search enabled. 50 answers. Answer share = fraction of answers naming the vendor. Named first = fraction where it was the first tracked vendor named.

**Hugging Face leads** with 96% answer share (named in 48 of 50 answers) and was named first in 34%. Together follows at 90%. 18 vendors were named at least once.

## Answer share

| # | Vendor | Answer share | Named first | Mentions | Models |
|---|---|---|---|---|---|
| 1 | Hugging Face | 96% | 34% | 48/50 | GPT-5.6 Sol, ChatGPT, GPT-5.6 Luna, Claude Opus 5, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash, Perplexity, Sonar Reasoning Pro |
| 2 | Together | 90% | 38% | 45/50 | GPT-5.6 Sol, ChatGPT, GPT-5.6 Luna, Claude Opus 5, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash, Perplexity, Sonar Reasoning Pro |
| 3 | Fireworks | 86% | 8% | 43/50 | GPT-5.6 Sol, ChatGPT, GPT-5.6 Luna, Claude Opus 5, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash, Perplexity, Sonar Reasoning Pro |
| 4 | Baseten | 72% | 0% | 36/50 | GPT-5.6 Sol, ChatGPT, GPT-5.6 Luna, Claude Opus 5, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash, Perplexity, Sonar Reasoning Pro |
| 5 | Modal | 68% | 4% | 34/50 | GPT-5.6 Sol, ChatGPT, GPT-5.6 Luna, Claude Opus 5, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash, Sonar Reasoning Pro |
| 6 | RunPod | 68% | 0% | 34/50 | GPT-5.6 Sol, GPT-5.6 Luna, Claude Opus 5, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash, Perplexity, Sonar Reasoning Pro |
| 7 | Groq | 64% | 10% | 32/50 | GPT-5.6 Sol, ChatGPT, GPT-5.6 Luna, Claude Opus 5, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash, Perplexity, Sonar Reasoning Pro |
| 8 | Replicate | 46% | 0% | 23/50 | GPT-5.6 Sol, ChatGPT, GPT-5.6 Luna, Claude Opus 5, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash, Perplexity, Sonar Reasoning Pro |
| 9 | Ollama | 42% | 2% | 21/50 | GPT-5.6 Luna, Claude Opus 5, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash, Perplexity, Sonar Reasoning Pro |
| 10 | Amazon Bedrock | 38% | 0% | 19/50 | Claude Opus 5, Claude, Claude Fable 5, Gemini, Perplexity, Sonar Reasoning Pro |
| 11 | SiliconFlow | 34% | 2% | 17/50 | Claude Opus 5, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash, Perplexity, Sonar Reasoning Pro |
| 12 | Lambda | 28% | 0% | 14/50 | GPT-5.6 Sol, GPT-5.6 Luna, Claude Opus 5, Claude, Gemini, Gemini 3.5 Flash |
| 13 | DeepInfra | 16% | 0% | 8/50 | GPT-5.6 Sol, Claude, Claude Fable 5, Gemini, Gemini 3.5 Flash |
| 14 | Vast.ai | 16% | 0% | 8/50 | Claude, Gemini, Gemini 3.5 Flash |
| 15 | OpenRouter | 12% | 2% | 6/50 | GPT-5.6 Sol, Claude Opus 5, Claude Fable 5, Gemini 3.5 Flash |
| 16 | Cerebras | 4% | 0% | 2/50 | Claude Opus 5, Claude Fable 5 |
| 17 | Anyscale | 2% | 0% | 1/50 | Gemini 3.5 Flash |
| 18 | Vercel AI | 2% | 0% | 1/50 | Gemini |

## By model

- **GPT-5.6 Sol** (OpenAI, gpt-5.6-sol, 5 answers): Modal 100%, Hugging Face 80%, Together 80%, Baseten 80%, Fireworks 60%
- **GPT-5.6 Terra** (OpenAI, gpt-5.6-terra, 5 answers): Hugging Face 80%, Together 60%, Fireworks 60%, Modal 60%, Baseten 40%
- **GPT-5.6 Luna** (OpenAI, gpt-5.6-luna, 5 answers): Hugging Face 100%, Modal 100%, RunPod 80%, Replicate 80%, Together 60%
- **Claude Opus 5** (Anthropic, claude-opus-5, 5 answers): Hugging Face 100%, Together 100%, Fireworks 100%, Baseten 100%, Groq 100%
- **Claude Sonnet 5** (Anthropic, claude-sonnet-5, 5 answers): Hugging Face 100%, Together 100%, Fireworks 100%, RunPod 80%, Groq 80%
- **Claude Fable 5** (Anthropic, claude-fable-5, 5 answers): Hugging Face 100%, Together 100%, Fireworks 100%, Baseten 100%, Groq 100%
- **Gemini 3.6 Flash** (Google, gemini-3.6-flash, 5 answers): Hugging Face 100%, Together 100%, Fireworks 100%, Modal 100%, RunPod 100%
- **Gemini 3.5 Flash** (Google, gemini-3.5-flash, 5 answers): Hugging Face 100%, Together 100%, Baseten 100%, Modal 100%, RunPod 100%
- **Sonar Pro** (Perplexity, sonar-pro, 5 answers): Hugging Face 100%, Together 100%, Fireworks 100%, RunPod 100%, Baseten 80%
- **Sonar Reasoning Pro** (Perplexity, sonar-reasoning-pro, 5 answers): Hugging Face 100%, Together 100%, Fireworks 100%, Groq 100%, RunPod 80%

## Sources the models cited

- llmapi.ai: 107
- edenai.co: 66
- siliconflow.com: 54
- databasemart.com: 53
- fireworks.ai: 48
- huggingface.co: 36
- spheron.network: 34
- hammansamuel.medium.com: 33
- reddit.com: 28
- checkthat.ai: 28
- runpod.io: 21
- thundercompute.com: 19
- northflank.com: 17
- qualixsolutions.com: 15
- guptadeepak.com: 14

## Prompts

1. What is the best platform to host and serve open-source models for a developer? Name specific products.
2. Which platform to host and serve open-source models would you recommend to a developer in 2026?
3. Compare the top platform to host and serve open-source models options right now.
4. I'm a developer and I need a platform to host and serve open-source models. What should I use and why?
5. Best platform to host and serve open-source models for fast, cheap inference?

## Tracked but never named

SambaNova

Raw answers: https://www.memetik.ai/index/inference-hosting (every answer published in full). Chart image: https://www.memetik.ai/charts/inference-hosting.png. Licence: CC BY 4.0, credit "Memetik Index".