aiapicostcalcUnit economicsPricesBrandsCompareBenchmarksSkills & PluginsFind my modelGuidesChanges

GPT-5.5 Rosalind Discovery pricing on the ChatGPT Enterprise rate card

Short answer: GPT-5.5 Rosalind Discovery costs $5 per 1M input tokens and $25 per 1M output tokens on OpenAI's ChatGPT Enterprise rate card (checked 2026-11-18), with cached input at $0.5. A 2,000-input / 500-output chat call costs $0.0225, so 1,000 calls cost $22.50. Output is 5.0× the input price, which is why long answers dominate a bill. OpenAI's API pricing page lists the Rosalind life-sciences models for approved organizations, with GPT-6 Rosalind at $10 / $50 and Rosalind Research and Discovery at $5 / $25 per 1M.

Last verified 2026-11-18 · Sources: OpenAI pricing page, checked 2026-11-18 · Report an error

By Shangyi Duan · Reviewed against OpenAI's ChatGPT Enterprise rate card

Source URL: https://help.openai.com/en/articles/20001415-chatgpt-rate-card-enterprise-token-based-pricing · Verified: · Rate row: src/prices.json

1
2
Presets

Preset token counts are starting assumptions; replace them with your logged averages.

Cost per call$0.0225
Per 1,000 calls$22.50

$5 in · $25 out per 1M

Nearest-priced models, this workload

Showing the 9 models closest in price; change any input to rank every model.

ModelPer callPer 1K calls
Claude Sonnet 4.6$0.0135$13.50
GPT-5.6 Sol$0.0180$18.00
Claude Opus 5.5$0.0180$18.00
GPT-5.5 Rosalind Research$0.0225$22.50
GPT-5.5 Rosalind Discovery$0.0225$22.50
Claude Opus 5$0.0225$22.50
GPT-5.5$0.0250$25.00
Claude Fable 5.2$0.0360$36.00
Claude Mythos 5.2$0.0360$36.00

o4-mini is shut down; its last published list rate is kept for comparing old invoices, not as a model you can call today — the model page names its replacement.

At a glance

API model IDgpt-rosalind-discovery
StatusSelect Enterprise plans only (Discovery approval)

Every price tier

TierPrice per 1M tokens
Standard input$5
Cached input (cache read)$0.5
Output (incl. reasoning tokens)$25
Batch API input / outputBatch not supported

Source: OpenAI ChatGPT Enterprise rate card, checked 2026-11-18.

ChatGPT Enterprise rate-card row for select Enterprise plans (Chat, Work and Codex), labelled GPT-5.5 Rosalind Discovery. Discovery approval does not grant Research access; the API pricing page now lists it under Life Sciences Trusted Access at $5/$25 per 1M short context (cached $0.50), with no long-context rate listed

Worked examples

WorkloadPer callPer 1K callsPer 1K, Batch
Chat reply · 1K in / 500 out$0.0175$17.50n/a
RAG answer · 10K in / 1K out$0.0750$75.00n/a
RAG with 80% cached prefix$0.0390$39.00n/a
Long document · 100K in / 2K out$0.5500$550n/a
Whole-codebase prompt · 400K in / 4K out$2.10$2,100n/a

Does prompt caching pay off?

There is no separate cache-write charge listed for this model, so every cache hit on a 20K-token prefix saves $0.0900 versus the normal input price ($90.00 per 1,000 calls). Cached reads cost 10% of the normal input price on this model.

When to pick it

Where it sits in the OpenAI lineup

Among the 21 OpenAI models we track, GPT-5.5 Rosalind Discovery ranks #16 from cheapest at a 10K-in / 1K-out workload ($75.00 per 1,000 calls). The next cheaper option is GPT-5.5 Rosalind Research at $75.00 (1× the cost). Stepping up to GPT-5.5 costs $80.00 (1.1×). The cheapest in the lineup, GPT-6 Luna, is 50× cheaper.

Which Rosalind row gets billed

ModelWhere it is billedInput / output per 1MAccess
GPT-6 RosalindAPI, ChatGPT Work and Codex$10 / $50 per 1MApproved organizations with Research access; not in the Chat tab
GPT-5.5 Rosalind ResearchChatGPT Chat, Work and Codex$5 / $25 per 1MSelect Enterprise plans
GPT-5.5 Rosalind DiscoveryChatGPT Chat, Work and Codex$5 / $25 per 1MSelect Enterprise plans; Discovery approval does not grant Research access

The $5 / $25 row on this page is the Enterprise Discovery one, priced on the GPT-5.5 generation. Discovery approval does not grant Research access, so a Discovery-only agreement cannot run GPT-6 Rosalind even though the API pricing page lists it for approved organizations.

Work that later needs GPT-6 Rosalind is billed at $10 input and $50 output per 1M tokens, in ChatGPT Work and Codex only. Budget the switch when the agreement adds Research access, and keep the Discovery row as the baseline until then.

Rates: OpenAI ChatGPT Enterprise rate card and OpenAI API pricing, both read on 2026-11-15.

Related

Ready to build? Create an API key in the official OpenAI console.

Questions

How much does GPT-5.5 Rosalind Discovery cost per 1M tokens?

$5 for input, $25 for output and $0.5 for cached input on the ChatGPT Enterprise rate card. Source: the official OpenAI ChatGPT Enterprise rate card, checked 2026-11-18.

How much do 1,000 GPT-5.5 Rosalind Discovery calls cost?

At 2,000 input and 500 output tokens per call, one call costs $0.0225 and 1,000 calls cost $22.50 at list price. Batch and caching are not included; both lower the total.

Is GPT-5.5 Rosalind Discovery expensive compared with other models?

On a 2,000-input / 500-output chat call it ranks #43 of 55 verified models by cost. The cheapest is Ministral 3 3B at $0.00025 per call.

Does GPT-5.5 Rosalind Discovery publish a cached-input rate or a batch discount?

Its ChatGPT Enterprise rate card lists a cached-input rate of $0.5 per 1M tokens and no Batch API rate. Caching pays off when a system prompt or document prefix repeats across calls; batch applies only to asynchronous jobs nobody is waiting on. Both are priced on top of the standard rates above, never instead of them.

Next step → Estimate cost per user with GPT-5.5 Rosalind Discovery

Price data last checked 2026-11-18 against official sources · how we verify

Price-change alerts

One short email when a provider changes a price or a promo ends. No spam, unsubscribe any time.

Tools, skills & plugins that use this model

Matched to this model's profile: cheaper models are paired with workflow and agent tools, million-token-context models with frameworks and open-source projects, and reasoning models with coding agents. Each entry links to its token-cost page or official site, and the per-task price is calculated on this model from the tool data rather than copied from another model.

More paths from this model