Inference.net logo

Inference.net

LLM 5.1/10 Jun 17, 2026

Serverless LLM inference, fine-tuning, evals, and gateway routing with a free-forever plan and $1/month credit grant.

Requests / month
1M
Free credit
$1
Credit window
30 days

Inference.net scores 5.1/10 for free value in LLM. The main constraint: Free-forever plan with a $1/month credit grant, 1 active model deployment (unlimited deployment GPU hours), 2 fine-tuning jobs and 500 eval samples per month, unlimited third-party serverless inference routing, 1M gateway inference requests with 50 GB data per month, and 1M OpenTelemetry spans with 50 GB span data per month.

Category LLM Kind LLM inference & fine-tuning platform Caveats The $1/month credit covers only a small amount of inference; serverless usage beyond it is pay-as-you-go and requires adding a valid payment card. Signal classifications (span/trace/session) are excluded on the free plan, and only 1 active deployment is allowed across all tiers. Renewal monthly Card required No Eligibility anyone

Free value score

5.1 / 10
Show the math
Value 5/10
value = 1 $/mo
weighted score = 0.7 * 5/10
Model Quality 4.3/10
value = 24 artificial_analysis_score
weighted score = 0.2 * 4.3/10
Volume 5.3/10
value = 13477089 tok/mo
score = 5.3/10
Requests 7/10
value = 1000000 req/mo
weighted score = 0.1 * 7/10
Access 7/10
value = 7/10
score = 7/10

Key metrics

gpt-oss-120b 24 Best model
1.0M R/mo Requests/month
$1 Credit
Yes Commercial use

Models

22 free · 22 total

Limits may be stale — last re-verified outside this section's freshness window.

Model Family AA Access
gpt-oss-120b GPT 23.8 free
DeepSeek R1 DeepSeek 20.1 free
DeepSeek R1 0528 DeepSeek 20.1 free
gpt-oss-20b GPT 14.9 free
DeepSeek V3 DeepSeek 10.4 free
DeepSeek V3 0324 DeepSeek 10.4 free
Gemma 3 27B Instruct Gemma 4.8 free
ClipTagger 12B Gemma — free
DeepSeek R1 Distill Llama 70B DeepSeek — free
Llama 3.1 70B Instruct Llama — free
Llama 3.1 8B Instruct Llama — free
Llama 3.2 11B Instruct Llama — free
Llama 3.2 1B Instruct Llama — free
Llama 3.2 3B Instruct Llama — free
Llama 3.3 70B Instruct Llama — free
Mistral Nemo 12B Instruct Mistral — free
Nemotron 3 Super Nemotron — free
Qwen2.5 7B Instruct Qwen — free
Qwen3 30B A3B Qwen — free
QwQ-32B Qwen — free
Schematron V2 Small Llama — free
Schematron V2 Turbo Granite — free

Compare all free models across providers →

What you get

$0 forever plan with a recurring $1/month credit grant (no card needed until you exceed it)

1 active model deployment with unlimited deployment GPU hours

1M gateway inference requests/mo (50 GB) plus 1M OpenTelemetry observability spans/mo (50 GB)

Unlimited third-party serverless inference routing and unlimited HALO runs/inference

Gotchas

  • 1 active deployment

    Only 1 active model deployment is allowed on the free plan (also capped at 1 on paid tiers).

  • 50 GB gateway data/mo

    Gateway inference data limited to 50 GB per month.

  • 50 GB OTEL data/mo

    OpenTelemetry span data limited to 50 GB per month.

  • Card needed for PAYG

    Serverless inference beyond the $1 credit is pay-as-you-go; per Terms of Service a valid payment card is required to use any non-free aspect of the service.

Sources

  • "Free $0 forever ... $1 monthly credit grant ... 1 deployment Active deployments ... Unlimited Deployment GPU hours ... 2 / mo Training jobs ... 500 / mo Eval samples ... Unlimited Third-party serverless inference ... 1M / mo Gateway inference requests ... 50 GB / mo Gateway inference data ... 1M / mo OTEL spans ... 50 GB / mo OTEL span data ... Pay-as-you-go serverless inference"

    https://inference.net/pricing/

  • "we have raised $11.8M in Series Seed funding, led by Multicoin Capital and a16z CSX"

    https://inference.net/blog/seed-round/

  • "Kimi K2 Thinking achieves a 67 in the [Artificial Analysis] Intelligence Index. This positions it clearly above all other open weights models, including the recently released MiniMax-M2 and DeepSeek-V3.2-Exp, and second only to GPT-5 amongst proprietary models."

    https://x.com/ArtificialAnlys/status/1986911675820446013

  • "Deploy any model in 5 minutes ... Kimi K2.5 $9.98/hr B200 · 180 GiB VRAM ... GLM-5 $9.98/hr ... MiniMax-M2.5 $9.98/hr ... GPT-OSS 120B $9.98/hr"

    https://inference.net/products/deploy/
llm inference fine-tuning serverless gateway observability open-models ai
Open service Source ← All LLM

Alternatives

Same category · similar quota
OpenCode GoOpenCode Go · paid · Paid plan
5.2

~$60 of usage/month (dollar limits: $12/5h, $30/week, $60/month) across open coding models incl. DeepSeek V4 Pro, GLM-5.2, Qwen3.7 Max, Kimi K2.7 Code, and MiniMax M3

1×44.3 AA1.0×
May be outdatedChecked 26d ago
Paid onlyMetered
OpenRouter CreditsOpenRouter Credits · paid · Paid plan
4.8

Pre-paid credits routed to 300+ models at provider list prices; credits never expire; Kimi K2 Thinking is $0.60/$2.50 per 1M on OpenRouter; 1M free BYOK requests/mo then 5% BYOK fee

930.2K/$19.4 AA930.2K/$
May be outdatedChecked 26d ago
Paid onlyEmbeddings
SiliconFlowSiliconFlow · freemium · Hosted open-weight inference (CN)
4.7

China-based inference platform offering three permanently free open-weight models alongside 200+ paid models.

15 AA30/min
Checked 10d ago