Baseten logo

Baseten

LLM 8.0/10

Usage-based platform for hosting open-weight LLMs and custom models, with $30 in free credits for new workspaces and no platform fee.

Baseten scores 8/10 for free value in LLM. The main constraint: New workspaces receive a one-time $30 in free credits (no credit card required at signup) that apply to Model APIs and dedicated/ training compute on the usage-based Startup plan.

Category LLM Kind LLM inference platform Caveats The $30 is a one-time credit, not a renewing monthly allowance, and burns down at usage-based rates (per-token for Model APIs, per-GPU-minute for dedicated deployments); once exhausted, a credit card and pay-as-you-go billing are required to continue. Renewal one_time Card required No Eligibility anyone

Free value score

8.0 / 10
Show the math
Value 8/10
value = 30 $ one-time
weighted score = 0.7 * 8/10
Model Quality 9.2/10
value = 51 artificial_analysis_score
weighted score = 0.2 * 9.2/10
Volume 5.4/10
value = 15584416 tok/mo
weighted score = 0.1 * 5.4/10
Access 7/10
value = 7/10
score = 7/10

Key metrics

GLM 5.2 51 Best model
$30 Credit
Yes Commercial use

Models

13 free · 87+ total
Model Family AA Access
GLM 5.2 GLM 51.1 free
DeepSeek V4 DeepSeek 44.3 free
Kimi K2.6 Kimi 42.8 free
Kimi K2.7 Code Kimi 42 free
GLM 5.1 GLM 40.2 free
GLM 5 GLM 39.5 free
Kimi K2.5 Kimi 38.1 free
NVIDIA Nemotron 3 Ultra Nemotron 37.8 free
GLM 4.7 GLM 33.8 free
GPT OSS 120B GPT 23.8 free
DeepSeek V3 0324 DeepSeek 10.4 free
NVIDIA Nemotron 3 Super Nemotron — free
Qwen3 Coder 480B Qwen — free
GLM 4.7 Flash GLM 33.8 paid
DeepSeek V3.2 DeepSeek 24.7 paid
GLM 4.6 GLM 23 paid
DeepSeek V3.1 DeepSeek 21.1 paid
DeepSeek R1 0528 DeepSeek 20.1 paid
Kimi K2 Thinking Kimi 19.4 paid
GLM-4.5 Air GLM 16.5 paid
GPT OSS 20B GPT 14.9 paid
Llama 4 Maverick Llama 14.3 paid
DeepSeek-R1 Distill Qwen 32B DeepSeek 11 paid
Llama 4 Scout Llama 10 paid
Llama 3.1 Nemotron Ultra 253B Nemotron 9.1 paid
DeepSeek Prover V2 671B DeepSeek — paid
DeepSeek-R1 Distill Llama 70B DeepSeek — paid
DeepSeek-R1 Zero DeepSeek — paid
GLM-4.5V GLM — paid
GLM-4.6V GLM — paid
Kimi K2 Instruct (0905) Kimi — paid
Llama 3.3 70B Instruct Llama — paid
Llama 3.3 Nemotron 49B Super Nemotron — paid
Nemotron 3 Nano Omni Nemotron — paid
NVIDIA Nemotron 3 Nano Nemotron — paid
Qwen3 235B 2507 Qwen — paid
Qwen3 Coder 30B Qwen — paid
Qwen3 Next 80B A3B (Instruct & Thinking) Qwen — paid
Qwen3 VL 235B Qwen — paid
Qwen3.5 / Qwen3.6 series (4B-122B) Qwen — paid

Showing first 13 of a larger catalog.

What you get

$30 in free credits for every new workspace, no credit card needed to sign up

OpenAI-compatible Model APIs for open-weight LLMs (DeepSeek, GLM, Kimi K2, GPT-OSS 120B, Qwen)

Purely usage-based pricing with no monthly or annual platform fee on the default Startup plan

Credits also cover dedicated GPU deployments (T4 through B200) and fine-tuning/training jobs

Limitations

  • $30 one-time credit cap

    Free usage is capped at $30 of credit, which burns down at usage-based rates (per-token for Model APIs, per-GPU-minute for dedicated deployments). There is no fixed token or request allowance.

  • Card required after credits run out

    No credit card is needed to sign up and use the $30, but a card and pay-as-you-go billing are required to continue once the credit is exhausted.

Sources

llm inference open-weight model-hosting gpu openai-compatible usage-based free-credits
Open service Source ← All LLM