whichdevtool?
Menu
StacksDirectory Pricing changes Glossary
VPS and cloud servers App hosting and PaaS Managed databases Transactional email APIs Email marketing and newsletters Domain registrars CDN, DNS and edge security Monitoring, uptime and error tracking Auth and identity Frontend and full-stack frameworks Backend frameworks UI component libraries AI model APIs File and object storage Payments and billing Background jobs, queues and workflows ORMs and database toolkits

Inference provider (open-weight models) · by SambaNova

SambaNova Cloud (SambaCloud)

Inference API running open-weight models such as DeepSeek, gpt-oss, Gemma, Llama and MiniMax on SambaNova's own RDU chips, with a small free tier.

Visit cloud.sambanova.ai ↗ Alternatives Free tier
SambaNova Cloud (SambaCloud) homepage
§1

Overview

SambaNova builds its own AI chips (RDUs) and SambaNova Cloud, also called SambaCloud, is its API for open-weight models running on them. The public catalogue is short and changes often; on 2026-09-29 it included gpt-oss-120b, DeepSeek V3.1 and V3.2, Gemma 4 31B, Llama 3.3 70B and MiniMax M2.7 and M3.

Accounts sit on the free tier until a payment method is linked. SambaNova's docs list free-tier limits of 20 requests a minute, 20 requests a day and 200,000 tokens a day for models such as DeepSeek-V3.2 and gemma-4-31B-it, which is enough for testing but not for an app. Linking a card moves the account to the Developer tier, billed per token with a ceiling of 20M tokens a day across all models.

Prices are mixed: gpt-oss-120b is $0.22 in and $0.59 out per million tokens, while DeepSeek-V3.2 is $3 / $4.50, well above DeepSeek's own API. Enterprise plans with higher limits are by quote.

§2

Pricing and plans

Free tier · no card

Free tier applies when no payment method is linked: for example DeepSeek-V3.2 and gemma-4-31B-it at 20 requests/min, 20 requests/day and 200,000 tokens/day.

At the limit Not stated The vendor does not say what happens at the limit, or we have not found it yet.

Plan Price What you get
Free tier $0 No payment method; 20 requests/day and 200K tokens/day per model
Developer tier Pay per token Payment method linked; up to 20M tokens/day across models
gpt-oss-120b $0.22 in / $0.59 out per 1M Developer tier
gemma-4-31B-it $0.38 in / $1.15 out per 1M Developer tier
MiniMax-M2.7 / M3 $0.60 in / $2.40 out per 1M Cached input $0.06 on M2.7
Meta-Llama-3.3-70B-Instruct $0.60 in / $1.20 out per 1M Developer tier
DeepSeek-V3.2 $3 in / $4.50 out per 1M Developer tier
Enterprise Custom Contact sales for higher limits

Per 1M tokens in USD, read 2026-09-29 on cloud.sambanova.ai/plans/pricing. Free-tier limits come from SambaNova's rate-limit docs (docs.sambanova.ai), found through search because the page blocked direct fetching; older copies of the page show different, higher limits, so check the console. SambaNova's plans page also says to add a payment method and purchase credits to run your first requests, which may mean the free tier now requires sign-up steps; this could not be resolved. A February 2025 SambaNova post described $5 of free credit for new developers, expiring after 3 months; whether this still applies was not confirmed.

Source: cloud.sambanova.ai/plans/pricing · prices as published, excluding tax.

§3

What it does

Custom RDU hardware
Models run on SambaNova's own chips rather than GPUs.
Free tier without a card
Per-model daily request and token caps before a payment method is linked.
Developer tier
Pay-per-token with up to 20M tokens a day across all models.
OpenAI-compatible API
Rate-limit headers report per-minute and per-day request limits.
Enterprise tier
Custom rate limits arranged with sales.
§4

Strengths and limitations

Strengths

  • Free testing without adding a card.
  • gpt-oss-120b at $0.22 / $0.59 per 1M tokens.
  • Developer tier daily cap bounds maximum daily spend.

Limitations

  • Free tier allows only 20 requests a day per model.
  • Developer tier is capped at 20M tokens a day across all models.
  • Short model catalogue that changes frequently.
  • Some models cost more than on their maker's own API (DeepSeek-V3.2 at $3 / $4.50).
§5

Who it suits

Good for

  • Benchmarking fast inference for open models
  • Small prototypes on gpt-oss or Llama

Look elsewhere if

  • High-volume production beyond 20M tokens a day without an enterprise deal
  • Apps needing a wide model catalogue
§6

Pricing history

  • SambaNova Cloud launches paid Developer tier

    Announced 2025-02-08: pay-per-token Developer tier with higher rate limits; existing and new developers received $5 of free credit expiring after 3 months. Source ↗

§7

Alternatives to SambaNova Cloud (SambaCloud)

Tool From Free option Stages
OpenRouter Model router (unified API)
— Free tier
Mistral API (Mistral Studio) AI model API (model maker)
Usage-based Free tier
OpenAI API AI model API (model maker)
Usage-based Paid
Together AI Inference provider (open-weight models)
Usage-based Paid
GroqCloud Inference provider (open-weight models)
Usage-based Free tier
Fireworks AI Inference provider (open-weight models)
Usage-based Trial

All AI model APIs →

§8

Questions

Is SambaNova Cloud free?

There is a free tier while no payment method is linked, limited per model (for example 20 requests a day and 200,000 tokens a day on DeepSeek-V3.2).

How do I get higher limits on SambaNova Cloud?

Link a payment method to move to the Developer tier, or contact sales for the Enterprise tier.

How much does gpt-oss-120b cost on SambaNova?

$0.22 per 1M input tokens and $0.59 per 1M output tokens on 2026-09-29.

Sources read for this page