> Markdown copy of https://perq.sh/blog/free-llm-api-tiers-november-update (Free LLM API tiers, November update | Perq), generated 2026-11-15. Index of every page for agents: https://perq.sh/llms.txt. Search API and MCP server: https://perq.sh/agents.
> Maritaca, SambaNova, Nscale and Alibaba Model Studio joined Perq today. Here is how they compare with the other free LLM API tiers we track, from each vendor's own page.

[Home](https://perq.sh/) › [Blog](https://perq.sh/blog)

# Free LLM API tiers, November update

By Perq team ([Avatar 8](https://perq.sh/about)) · Nov 12, 2026 · comparison · llm api · free tiers · inference

Offer details below are pulled from our dataset when the page is built, so they match the directory. Latest verification of offers in this post: Nov 15, 2026, 5:45 AM ET.

Four LLM API offers went public on Perq today: Maritaca AI’s S5-free model, SambaNova’s SambaCloud Free Tier, Nscale’s signup credit and Alibaba Cloud Model Studio’s free quota. Together with Inception and NVIDIA, which joined yesterday, that makes six new ways to call a hosted model without paying in the last two days. They work in different ways, so here is how they line up with the free tiers we already list. Every figure comes from the vendor’s own page.

| Vendor | Offer (vendor wording) | Eligibility | Verified | Links |
| --- | --- | --- | --- | --- |
| [Maritaca AI](https://perq.sh/offers/maritaca-s5-free) | [S5-free model: tokens not billed for a limited time (text only)](https://perq.sh/offers/maritaca-s5-free) | Anyone with a Maritaca AI API account | Nov 15, 2026, 5:45 AM ET | [source ↗](https://docs.maritaca.ai/en/precos) · [claim ↗](https://docs.maritaca.ai/en/precos) |
| [SambaNova](https://perq.sh/offers/sambanova-free-tier) | [SambaCloud Free Tier: rate-limited API access with daily token limits, no payment method needed](https://perq.sh/offers/sambanova-free-tier) | Anyone with a SambaCloud account and no payment method linked | Nov 15, 2026, 5:45 AM ET | [source ↗](https://docs.sambanova.ai/docs/en/models/rate-limits) · [claim ↗](https://docs.sambanova.ai/docs/en/models/rate-limits) |
| [Nscale](https://perq.sh/offers/nscale-free-credit) | [$5 free credit for serverless inference](https://perq.sh/offers/nscale-free-credit) | New Nscale Console accounts | Nov 15, 2026, 5:45 AM ET | [source ↗](https://www.nscale.com/blog/what-is-serverless-inference) · [claim ↗](https://www.nscale.com/blog/what-is-serverless-inference) |
| [Alibaba Cloud](https://perq.sh/offers/alibaba-model-studio-free-quota) | [Free token quota for many models (Qwen and others) when you first activate Model Studio, valid 90 days](https://perq.sh/offers/alibaba-model-studio-free-quota) | New Alibaba Cloud Model Studio users (Singapore region, International scope) with completed account information | Nov 15, 2026, 5:45 AM ET | [source ↗](https://www.alibabacloud.com/help/en/model-studio/new-free-quota) · [claim ↗](https://www.alibabacloud.com/help/en/model-studio/new-free-quota) |
| [Inception](https://perq.sh/offers/inception-mercury-free-tokens) | [100 million free tokens with every new Inception API key (Mercury models)](https://perq.sh/offers/inception-mercury-free-tokens) | Anyone who creates a new Inception API key | Nov 15, 2026, 5:45 AM ET | [source ↗](https://www.inceptionlabs.ai/blog/mercury-2-10x-free-tokens) · [claim ↗](https://www.inceptionlabs.ai/blog/mercury-2-10x-free-tokens) |
| [NVIDIA](https://perq.sh/offers/nvidia-build-free-endpoints) | [Free prototyping on build.nvidia.com model endpoints, up to 40 requests per minute for most models, no per-token billing](https://perq.sh/offers/nvidia-build-free-endpoints) | Anyone with an NVIDIA developer account | Nov 15, 2026, 5:45 AM ET | [source ↗](https://build.nvidia.com/api/runtime/content/help-center-faq.yaml) · [claim ↗](https://build.nvidia.com) |

## Three kinds of “free”

**Rate-limited free tiers (no fixed pot, limits reset).**

-   **SambaNova:** the Free Tier applies when no payment method is linked to your account. Limits are per model (requests per minute, requests per day, tokens per day), so there is no single number to quote.
-   **NVIDIA build.nvidia.com:** free prototyping, up to 40 requests per minute for most models, with no per-token billing.
-   **Groq, Gemini API and OpenRouter** (already listed) work the same way: free access capped by per-model or per-day rate limits.

**A one-time token or credit pot.**

-   **Inception:** every new API key includes 100 million free tokens for the Mercury models.
-   **Nscale:** a $5 free credit when you create a Nscale Console account. The only source is Nscale’s own blog post, so check the console when you sign up.
-   **Alibaba Cloud Model Studio:** a free token quota for many models (Qwen and others) when you first activate Model Studio. The quota differs by model and is valid for 90 days. After it runs out, usage is billed pay-as-you-go unless you turn on the “Free Quota Only” setting.
-   **Cerebras** (already listed): a one-time $5 promotional credit, but only after you add a payment method, and it expires 30 days after activation.

**Free while a promotion or experiment lasts.**

-   **Maritaca AI:** the S5-free model’s tokens are not billed for a limited time. It is text only, and Batch API, Flex and built-in tools are not included. No end date is stated, and the page says it may be discontinued at any time.
-   **Hetzner** (already listed): the Inference API is free while it stays experimental.

| Vendor | Offer (vendor wording) | Eligibility | Verified | Links |
| --- | --- | --- | --- | --- |
| [Groq](https://perq.sh/offers/groq-free-plan) | [Free Plan with per-model rate limits (RPM/RPD/TPM/TPD)](https://perq.sh/offers/groq-free-plan) | Anyone with a GroqCloud account | Nov 15, 2026, 5:45 AM ET | [source ↗](https://console.groq.com/docs/rate-limits) · [claim ↗](https://console.groq.com/) |
| [Google (Gemini API)](https://perq.sh/offers/gemini-api-free-tier) | [Free of charge input/output on Free Tier models (rate-limited)](https://perq.sh/offers/gemini-api-free-tier) | Anyone with an AI Studio API key (free tier project) | Nov 15, 2026, 5:45 AM ET | [source ↗](https://ai.google.dev/gemini-api/docs/pricing) · [claim ↗](https://aistudio.google.com/apikey) |
| [OpenRouter](https://perq.sh/offers/openrouter-free-models) | [Free :free model variants — 50 requests/day (1,000/day after buying 10 credits), 20 RPM](https://perq.sh/offers/openrouter-free-models) | Any OpenRouter account | Nov 15, 2026, 5:45 AM ET | [source ↗](https://openrouter.ai/pricing) · [claim ↗](https://openrouter.ai/openrouter/free) |
| [Cerebras](https://perq.sh/offers/cerebras-promo-credit) | [One-time $5 promotional credit (expires 30 days after activation)](https://perq.sh/offers/cerebras-promo-credit) | Accounts that add a valid payment method | Nov 15, 2026, 5:45 AM ET | [source ↗](https://www.cerebras.ai/pricing) · [claim ↗](https://cloud.cerebras.ai/) |
| [Hetzner](https://perq.sh/offers/hetzner-inference-api-free) | [Inference API free of charge while it is in experimental status](https://perq.sh/offers/hetzner-inference-api-free) | Hetzner customers | Nov 15, 2026, 5:45 AM ET | [source ↗](https://docs.hetzner.com/general/company-and-policy/experiments/inference/) · [claim ↗](https://docs.hetzner.com/general/company-and-policy/experiments/inference/) |

## Which one to start with

-   **An agent that runs every day:** a rate-limited tier (SambaNova, NVIDIA, Groq, Gemini) keeps working as long as you stay under the limits.
-   **A one-off batch or eval:** a fixed pot (Inception’s tokens, Alibaba’s quota) lets you run a larger job once without watching per-minute caps.
-   **Portuguese-language work:** Maritaca’s S5-free model is a Portuguese-first option, but plan for the promotion to end.

Rate limits, quotas and model lists change often. Check the linked source page before you build around a number, and turn on any “free quota only” or spending cap setting a vendor offers so a free trial can’t become a bill.

Perq pack

[LLM API free tier pack](https://perq.sh/packs/llm-api-free-tier-pack)

Model and embedding APIs you can call for free today: free tiers, daily quotas and small signup credits.

27 offers · 19 free tiers and 8 signup credits, no application needed

[See the pack →](https://perq.sh/packs/llm-api-free-tier-pack)

## Also public today

Three more offers came out of embargo today. Subscribers got them yesterday:

| Vendor | Offer (vendor wording) | Eligibility | Verified | Links |
| --- | --- | --- | --- | --- |
| [Windsurf](https://perq.sh/offers/windsurf-free) | [Free plan: $0, light agent quota, unlimited inline edits and Tab completions](https://perq.sh/offers/windsurf-free) | Anyone | Nov 15, 2026, 5:45 AM ET | [source ↗](https://windsurf.com/pricing) · [claim ↗](https://windsurf.com/pricing) |
| [Greptile](https://perq.sh/offers/greptile-starter-free) | [Starter plan: free, 50 AI code review credits per month for 1 active developer](https://perq.sh/offers/greptile-starter-free) | Individual developers | Nov 15, 2026, 5:45 AM ET | [source ↗](https://www.greptile.com/pricing) · [claim ↗](https://www.greptile.com/pricing) |
| [Sail Research](https://perq.sh/offers/sail-startups) | [Up to $10,000 in free credits for open-model inference and Sailboxes (persistent agent computers)](https://perq.sh/offers/sail-startups) | Bootstrapped and funded startups building with Sail; applications reviewed individually | Nov 15, 2026, 5:45 AM ET | [source ↗](https://www.sailresearch.com/startups) · [claim ↗](https://www.sailresearch.com/startups) |

Windsurf’s Free plan includes a light agent quota plus unlimited inline edits and Tab completions. Greptile’s free Starter plan gives one developer 50 AI code review credits a month. It replaces Greptile’s startup discount, which we marked closed yesterday. Sail Research offers startups up to $10,000 in credits for open-model inference and persistent agent computers. You have to apply, and approval isn’t guaranteed.

Want new offers a day before they show up on the site? [Join early access](https://perq.sh/newsletter).

## All offers mentioned in this post

| Vendor | Offer | Category | Type | Eligibility | Source | Last verified |
| --- | --- | --- | --- | --- | --- | --- |
| [Maritaca AI](https://perq.sh/offers/maritaca-s5-free) | [S5-free model: tokens not billed for a limited time (text only)](https://perq.sh/offers/maritaca-s5-free) | [LLM APIs](https://perq.sh/category/llm-api) | [free tier](https://perq.sh/type/free-tier) | Anyone with a Maritaca AI API account | [docs.maritaca.ai ↗](https://docs.maritaca.ai/en/precos) | Nov 15, 2026, 5:45 AM ET |
| [SambaNova](https://perq.sh/offers/sambanova-free-tier) | [SambaCloud Free Tier: rate-limited API access with daily token limits, no payment method needed](https://perq.sh/offers/sambanova-free-tier) | [LLM APIs](https://perq.sh/category/llm-api) | [free tier](https://perq.sh/type/free-tier) | Anyone with a SambaCloud account and no payment method linked | [docs.sambanova.ai ↗](https://docs.sambanova.ai/docs/en/models/rate-limits) | Nov 15, 2026, 5:45 AM ET |
| [Nscale](https://perq.sh/offers/nscale-free-credit) | [$5 free credit for serverless inference](https://perq.sh/offers/nscale-free-credit) | [LLM APIs](https://perq.sh/category/llm-api) | [signup credit](https://perq.sh/type/signup-credit) | New Nscale Console accounts | [nscale.com ↗](https://www.nscale.com/blog/what-is-serverless-inference) | Nov 15, 2026, 5:45 AM ET |
| [Alibaba Cloud](https://perq.sh/offers/alibaba-model-studio-free-quota) | [Free token quota for many models (Qwen and others) when you first activate Model Studio, valid 90 days](https://perq.sh/offers/alibaba-model-studio-free-quota) | [LLM APIs](https://perq.sh/category/llm-api) | [signup credit](https://perq.sh/type/signup-credit) | New Alibaba Cloud Model Studio users (Singapore region, International scope) with completed account information | [alibabacloud.com ↗](https://www.alibabacloud.com/help/en/model-studio/new-free-quota) | Nov 15, 2026, 5:45 AM ET |
| [Inception](https://perq.sh/offers/inception-mercury-free-tokens) | [100 million free tokens with every new Inception API key (Mercury models)](https://perq.sh/offers/inception-mercury-free-tokens) | [LLM APIs](https://perq.sh/category/llm-api) | [signup credit](https://perq.sh/type/signup-credit) | Anyone who creates a new Inception API key | [inceptionlabs.ai ↗](https://www.inceptionlabs.ai/blog/mercury-2-10x-free-tokens) | Nov 15, 2026, 5:45 AM ET |
| [NVIDIA](https://perq.sh/offers/nvidia-build-free-endpoints) | [Free prototyping on build.nvidia.com model endpoints, up to 40 requests per minute for most models, no per-token billing](https://perq.sh/offers/nvidia-build-free-endpoints) | [LLM APIs](https://perq.sh/category/llm-api) | [free tier](https://perq.sh/type/free-tier) | Anyone with an NVIDIA developer account | [build.nvidia.com ↗](https://build.nvidia.com/api/runtime/content/help-center-faq.yaml) | Nov 15, 2026, 5:45 AM ET |
| [Groq](https://perq.sh/offers/groq-free-plan) | [Free Plan with per-model rate limits (RPM/RPD/TPM/TPD)](https://perq.sh/offers/groq-free-plan) | [LLM APIs](https://perq.sh/category/llm-api) | [free tier](https://perq.sh/type/free-tier) | Anyone with a GroqCloud account | [console.groq.com ↗](https://console.groq.com/docs/rate-limits) | Nov 15, 2026, 5:45 AM ET |
| [Google (Gemini API)](https://perq.sh/offers/gemini-api-free-tier) | [Free of charge input/output on Free Tier models (rate-limited)](https://perq.sh/offers/gemini-api-free-tier) | [LLM APIs](https://perq.sh/category/llm-api) | [free tier](https://perq.sh/type/free-tier) | Anyone with an AI Studio API key (free tier project) | [ai.google.dev ↗](https://ai.google.dev/gemini-api/docs/pricing) | Nov 15, 2026, 5:45 AM ET |
| [OpenRouter](https://perq.sh/offers/openrouter-free-models) | [Free :free model variants — 50 requests/day (1,000/day after buying 10 credits), 20 RPM](https://perq.sh/offers/openrouter-free-models) | [LLM APIs](https://perq.sh/category/llm-api) | [free tier](https://perq.sh/type/free-tier) | Any OpenRouter account | [openrouter.ai ↗](https://openrouter.ai/pricing) | Nov 15, 2026, 5:45 AM ET |
| [Cerebras](https://perq.sh/offers/cerebras-promo-credit) | [One-time $5 promotional credit (expires 30 days after activation)](https://perq.sh/offers/cerebras-promo-credit) | [LLM APIs](https://perq.sh/category/llm-api) | [signup credit](https://perq.sh/type/signup-credit) | Accounts that add a valid payment method | [cerebras.ai ↗](https://www.cerebras.ai/pricing) | Nov 15, 2026, 5:45 AM ET |
| [Hetzner](https://perq.sh/offers/hetzner-inference-api-free) | [Inference API free of charge while it is in experimental status](https://perq.sh/offers/hetzner-inference-api-free) | [LLM APIs](https://perq.sh/category/llm-api) | [free tier](https://perq.sh/type/free-tier) | Hetzner customers | [docs.hetzner.com ↗](https://docs.hetzner.com/general/company-and-policy/experiments/inference/) | Nov 15, 2026, 5:45 AM ET |
| [Windsurf](https://perq.sh/offers/windsurf-free) | [Free plan: $0, light agent quota, unlimited inline edits and Tab completions](https://perq.sh/offers/windsurf-free) | [Developer tools](https://perq.sh/category/dev-tools) | [free tier](https://perq.sh/type/free-tier) | Anyone | [windsurf.com ↗](https://windsurf.com/pricing) | Nov 15, 2026, 5:45 AM ET |
| [Greptile](https://perq.sh/offers/greptile-starter-free) | [Starter plan: free, 50 AI code review credits per month for 1 active developer](https://perq.sh/offers/greptile-starter-free) | [Developer tools](https://perq.sh/category/dev-tools) | [free tier](https://perq.sh/type/free-tier) | Individual developers | [greptile.com ↗](https://www.greptile.com/pricing) | Nov 15, 2026, 5:45 AM ET |
| [Sail Research](https://perq.sh/offers/sail-startups) | [Up to $10,000 in free credits for open-model inference and Sailboxes (persistent agent computers)](https://perq.sh/offers/sail-startups) | [LLM APIs](https://perq.sh/category/llm-api) | [startup program](https://perq.sh/type/startup-program) | Bootstrapped and funded startups building with Sail; applications reviewed individually | [sailresearch.com ↗](https://www.sailresearch.com/startups) | Nov 15, 2026, 5:45 AM ET |

No offers match these filters.

## More from the blog

-   [OpenAI free credits and startup offers, explained](https://perq.sh/blog/openai-free-credits)
-   [This week in AI credits: 30 new offers, one closed, one back (Nov 14, 2026)](https://perq.sh/blog/this-week-in-ai-credits-nov-14-2026)
-   [Free tiers for running agents: sandboxes, remote browsers and durable jobs](https://perq.sh/blog/free-agent-sandboxes-and-runtimes)

---
Published by Avatar 8 (Avatar 8 LLC), ops@avatar33.com. Data: https://perq.sh/api/offers.json · OpenAPI: https://perq.sh/openapi.json
