Novita AI
novita.ai · 2 records in Compute, Inference · JSON
Records 2
- caveatNovita AI Agent Sandbox
Computefree tieridentity emailno cardlimit volume-quotaevidence documentedstale on 2026-12-04
- caveatNovita LLM API
Inferencefree tieridentity emailno cardlimit request-rateevidence documentedstale on 2026-12-04
Novita AI Agent Sandbox
Compute- Free allowance
- $100 of promotional credits after an account setup survey, no card; a running sandbox carries 20 GB of ephemeral storage and 60 GB of persistent-storage allowance
- Verdict
- caveatusable, with a named limit or an unprobed read path
- Cost
- free tier
- Identity
- emailno card, no phone
- Binding limit
- volume-quota$100 of promotional Sandbox credits, one time, valid for 90 days
- Evidence
- documented19 sources; a first-party page was read and the sentence quoted
- Freshness
- stale on 2026-12-04checked 2026-11-03 (live); stale on 2026-12-04
| Free tier | $100 of promotional credits after an account setup survey, no card; a running sandbox carries 20 GB of ephemeral storage and 60 GB of persistent-storage allowance unit mixed; metered on wall-clock |
|---|---|
| Price | none stated |
| Limits |
not stated |
| Binding | volume-quota the one-time $100 Sandbox credit, valid for 90 days, runs out before any other stated limit |
| Identity | email 'No credit card is required' for the promotional credits, quoted from the sandbox pricing page read 2026-10-22; the survey is the second gate |
| Egress | not stated |
| Per-object cap | not stated |
| Retention | not stated |
| Interface | unknown |
| Reliability | no reliability event recorded |
| Notes | The Novita LLM API model list prices apodex/apodex-1.1-mini at 0 per million input and output tokens and sets its default per-model limits at tier T1 to 30 RPM and 50000000 TPM; T2 is 100 RPM and T5 is 6000 RPM, with TPM at 50000000 in every tier. |
| Evidence note | documented the sandbox pricing page was read on 2026-10-22. The $100 is one-time and conditional on a survey, so the record states the condition rather than presenting it as a tier |
| Parsed numbers |
|
Evidence 19
| date | source | what it proved |
|---|---|---|
| 2026-10-02 | docs.novita.ai/guides/sandbox-pricing | not published |
| 2026-10-22 | docs.novita.ai/guides/sandbox-pricing | 'New users can complete a quick account setup survey to receive $100 in promotional credits for eligible Novita Sandbox usage. No credit card is required'; 'Ephemeral Storage: Each sandbox includes 20 GB ... while running'; paused sandboxes stop CPU and RAM billing and are charged storage only', quoted |
| 2026-10-22 | boat.dev/compare | a competitor's comparison table lists Novita at USD 0.233 per hour for a 4 vCPU / 8 GB / 50 GB machine, with list prices dated 2026-09-26; this is a reported rate about a third party and not a reading of the vendor's own page |
| 2026-10-30 | api-server.novita.ai/v1/product/model/list | the LLM API model list prices apodex/apodex-1.1-mini at 0 input and 0 output per million tokens and gives it 30 RPM and 50000000 TPM at tier T1 |
| 2026-10-30 | docs.novita.ai/guides/sandbox-pricing | the $100 Sandbox credit is one-time for new users, conditioned on an account setup survey, with no card |
| 2026-10-30 | novita.ai/user/register | the sign-up page grants $100 in Sandbox credits that expire after 90 days |
| 2026-10-30 | docs.novita.ai/guides/sandbox-pricing | a running sandbox has 20 GB of ephemeral storage and each account has 60 GB of free persistent storage |
| 2026-10-30 | docs.novita.ai/guides/quickstart | an account can be created with an email address alone, besides Google or GitHub sign-in |
| 2026-10-30 | docs.novita.ai/guides/llm-rate-limits | LLM API tier T1 is an account whose monthly top-ups did not exceed $50 in any of the last 3 calendar months, so an account with no top-up is T1 |
| 2026-10-30 | docs.novita.ai/guides/llm-rate-limits | LLM API RPM and TPM limits are counted per model and per account |
| 2026-10-30 | docs.novita.ai/_mintlify/custom/novitaai/64fac5c27e7ad00f/sc | the rate-limit page's per-tier RPM/TPM table is filled in the browser from https://api-server.novita.ai/v1/product/model/list |
| 2026-10-30 | novita.ai/pricing | the pricing page lists Apodex 1.1 Mini with 256K context at Free input and Free output |
| 2026-10-30 | blogs.novita.ai/apodex-1-1-mini-on-novita-ai-api-quickstart- | Novita describes the $0 price of Apodex 1.1 Mini, listed from 2026-10-01, as a launch promotion |
| 2026-10-30 | www.linkedin.com/posts/apodex_apodex-11-mini-api-playground- | Novita's launch post announced Apodex 1.1 Mini as free for 7 days; the $0 price is open to withdrawal |
| 2026-10-30 | docs.novita.ai/guides/error | the LLM API returns HTTP 403 for some models until the account completes identity verification; the page does not name the models |
| 2026-10-30 | docs.novita.ai/guides/LLM-FAQ | insufficient account balance is a documented cause of HTTP 403 on an LLM API call; the page does not say whether a $0 model is exempt |
| 2026-10-30 | docs.novita.ai/guides/LLM-FAQ | first-time accounts receive a New User Voucher for model services; the page states no amount |
| 2026-11-03 | api-server.novita.ai/v1/product/model/list | the LLM API model list prices apodex/apodex-1.1-mini at 0 output per million tokens and gives it 30 RPM at T1 and 100 RPM at T2, with 50000000 TPM |
| 2026-11-03 | novita.ai/pricing | the pricing page lists Apodex 1.1 Mini at Free input and Free output |
More fields 8
| keep rate | 1.0 |
|---|---|
| billing model | unclassified |
| session cap | {"wall_clock_min": null,"idle_reap_s": null,"restart_required": false,"note": "a paused sandbox stops CPU and RAM billing and keeps its data as persistent storage"} |
| persistent storage | {"included_gib": 60,"price_gib_month": "billed daily past the free allowance","durability": "paused sandboxes are retained as persistent storage","survives_stop": true,"survives_destroy": false,"max_retention": "until deleted"} |
| free tier kind | one-time-credit |
| granularity | {"min_billed_s": 0,"rounding_s": null} |
| burst x | 1.0 |
| crosscheck | {"boat_compare_2026-09-26": "USD 0.233/h at 4 vCPU / 8 GB"} |
Unknown values 10
- chaining: a source was read and is silent on this field
- egress metered: egress metering was not published
- egress terms: a source was read and is silent on this field
- interface: the access interface was not established
- limits note: a source was read and is silent on this field
- per object cap stated: a source was read and is silent on this field
- reliability note: nothing of this kind was observed
- retention notes: the field may be stated in a source that was not read
- retention stated: a source was read and is silent on this field
- tooling: a source was read and is silent on this field
Novita LLM API
Inference- Free allowance
- Apodex/apodex-1.1-mini (256K context) is priced at $0 per million input tokens and $0 per million output tokens on the serverless LLM API; no token quota is stated, and the model is capped by the default per-model limit of 30 requests per minute and 50000000 tokens per minute at tier T1.
- Verdict
- caveatusable, with a named limit or an unprobed read path
- Cost
- free tier
- Identity
- emailno card, no phone
- Binding limit
- request-rateapodex/apodex-1.1-mini default per-model limit by account tier: T1 30 requests per minute, T2 100, T3 1000, T4 3000, T5 6000; 50000000 tokens per minute at every tier; T1 is an account whose monthly top-ups did not exceed $50 in any of...
- Evidence
- documented17 sources; a first-party page was read and the sentence quoted
- Freshness
- stale on 2026-12-04checked 2026-11-03 (live); stale on 2026-12-04
| Free tier | apodex/apodex-1.1-mini (256K context) is priced at $0 per million input tokens and $0 per million output tokens on the serverless LLM API; no token quota is stated, and the model is capped by the default per-model limit of 30 requests per minute and 50000000 tokens per minute at tier T1. The $0 price was listed from 2026-10-01 as a launch promotion and was still listed on 2026-11-03 unit requests; metered on request-count; the unit is the nearest term: the allowance is a rate limit on a $0-priced model, with no token or credit quota; the key-less model list also prices the chat ids bunny, nvrouter/auto, nvrouter/fusion and dev/glm46 at 0, and no page read documents them or their limits, so only apodex/apodex-1.1-mini is counted; the $100 Sandbox credit is a separate product, in compute/novita |
|---|---|
| Price | none stated |
| Limits |
the rate limits are the vendor's defaults read from its model list; enforcement is not measured |
| Binding | request-rate the $0 model is bound by its per-model limit of 30 requests per minute at tier T1; its price is a promotion with no stated end |
| Identity | email an account created with an email address, Google, GitHub or Hugging Face sign-in; the sign-up page shows no phone or card field. unverified: whether a $0 model call needs a non-zero balance or a top-up, since the FAQ lists insufficient balance as a cause of HTTP 403, and whether apodex/apodex-1.1-mini needs identity verification, since some models do; both are behind a sign-up |
| Egress | not published |
| Interface | rest OpenAI-compatible chat/completions, plus Anthropic-style and Responses endpoints listed for the model |
| Tooling | any OpenAI-compatible client with base URL https://api.novita.ai/openai and model id apodex/apodex-1.1-mini; the model list GET /openai/v1/models answers without a key |
| Composes with | one key reaches every Novita-hosted model, so the $0 model and paid models share a key; a single $0 model is the only zero-priced model that a page documents |
| Reliability | the $0 price is a promotion with no stated end date, so the model can turn paid without notice; some models answer HTTP 403 until identity verification |
| Notes | re-check the price on a short interval: it was announced as free for 7 days on 2026-10-01 and was still $0 on 2026-11-03. Same vendor as compute/novita (Agent Sandbox), a different product with its own credit. |
| Evidence note | documented the $0 price, the tier limits and the sign-up options are quoted from vendor pages and the vendor's key-less model list, which a probe read on 2026-11-03; calling the model with a key is not measured |
| Parsed numbers |
|
Evidence 17
| date | source | what it proved |
|---|---|---|
| 2026-11-03 | api-server.novita.ai/v1/product/model/list | the product model list prices apodex/apodex-1.1-mini at 0 for input and output tokens |
| 2026-11-03 | api-server.novita.ai/v1/product/model/list | apodex/apodex-1.1-mini has 30 RPM and 50000000 TPM at tier T1 and 100 RPM at T2 |
| 2026-11-03 | api.novita.ai/openai/v1/models | the key-less OpenAI-compatible model list of the endpoint prices apodex/apodex-1.1-mini at 0 for input and output |
| 2026-11-03 | api.novita.ai/openai/v1/models | the same list prices a chat model id bunny at 0 too, with no page that documents it |
| 2026-11-03 | probe: api.novita.ai/openai/v1/models | a key-less GET of the OpenAI-compatible model list answered HTTP 200 on 2026-11-03 |
| 2026-11-03 | novita.ai/pricing | the public pricing page lists Apodex 1.1 Mini with 256K context at Free input and Free output |
| 2026-11-03 | novita.ai/models/model-detail/apodex-apodex-1.1-mini | the model page offers the model on the serverless pay-per-token API, with the input and output price shown as $0 |
| 2026-11-03 | novita.ai/models/model-detail/apodex-apodex-1.1-mini | the model page prices output at $0 per million tokens |
| 2026-11-03 | blogs.novita.ai/apodex-1-1-mini-on-novita-ai-api-quickstart- | the $0 price is a launch promotion listed from 2026-10-01 and the vendor tells readers to check the live price |
| 2026-11-03 | docs.novita.ai/guides/llm-rate-limits | tier T1 is the account tier with top-ups of at most $50, so an account with no top-up is T1 |
| 2026-11-03 | docs.novita.ai/guides/llm-rate-limits | limits are counted per account and per model |
| 2026-11-03 | docs.novita.ai/guides/quickstart | an account is created with Google, GitHub or an email address |
| 2026-11-03 | novita.ai/user/register | the sign-up page offers an email form besides the Google, GitHub and Hugging Face buttons and shows no phone or card field |
| 2026-11-03 | docs.novita.ai/guides/quickstart | the quickstart says credit and model access can decide whether an LLM request succeeds, without saying whether a $0 model needs a balance |
| 2026-11-03 | docs.novita.ai/guides/error | some models answer HTTP 403 until the account completes identity verification; the page names no model |
| 2026-11-03 | docs.novita.ai/guides/LLM-FAQ | insufficient account balance is a listed cause of an HTTP 403 on a model call; the page does not say whether a $0 model is exempt |
| 2026-11-03 | docs.novita.ai/guides/LLM-FAQ | first-time registered users get a voucher for model services; no amount is stated |
Unknown values 3
- egress metered: egress metering was not published
- per object cap stated: the field does not apply to this kind of offer
- retention stated: the field does not apply to this kind of offer