GPT-5.5 — Spécifications
| Développeur | OpenAI |
|---|---|
| Type | LLM (raisonnement) |
| Modalité | Texte, vision → texte |
| Paramètres | Non divulgué |
| Fenêtre de contexte | 1,05 million |
| Sortie maximale | 128 K |
| Licence | Propriétaire |
| Poids ouverts | Non |
| Publié | 2026 |
| Prix de l’entrée | 5,00 $ / 1 million |
| Prix de la sortie | 30,00 $ / 1 million |
| Fournisseurs d'API | OpenAI, Azure |
What is GPT-5.5?
GPT-5.5 is OpenAI’s flagship, covering frontier reasoning across coding, writing, research
and multimodal tasks with a context window of roughly 1.05M tokens. Pricing is $5 in / $30
out per million tokens, and a GPT-5.5 Pro tier exists for the highest-stakes reasoning.
Two pricing details shape how it should be used. First, prompts above 272K input tokens
are billed at a higher rate, so the effective cost of a long-context feature is not the
headline number — budget the premium explicitly or keep prompts under the threshold with
tighter retrieval. Second, the 6× output-to-input ratio is one of the steepest in the
frontier bracket: at $30 per million output tokens, generation-heavy workloads such as long
report writing or verbose agent traces cost far more than prompt-heavy ones like retrieval
and classification. Teams that instrument their token mix usually find output is where the
bill actually lives, and that trimming verbosity, capping max_tokens and moving first drafts
to a cheaper model saves more than any prompt-compression trick. Compare it directly against
Claude Opus 4.8, which charges $25 output with no long-context premium.
GPT-5.5 pricing: API cost per 1M tokens
| Entrée (par million de jetons) | $5.00 |
|---|---|
| Sortie (par million de jetons) | $30.00 |
| Output/input ratio | 6× |
| Blended (4:1 in:out) | $10.00 per 1M tokens |
What GPT-5.5 costs per month
Real monthly spend at a 4:1 input-to-output mix — the ratio a typical chat or RAG workload actually produces.
| Charge de travail | Jetons/mois | Cost / month |
|---|---|---|
| Projet secondaire | 1 million en entrée / 0,25 million en sortie | $13 |
| Petite équipe | 20 millions en entrée / 5 millions en sortie | $250 |
| Production | 200 millions en entrée / 50 millions en sortie | $2,500 |
Run your own numbers in the Calculateur de coûts des API IA.
Cheaper alternatives to GPT-5.5
| Modèle | Coût combiné par million de dollars | You save |
|---|---|---|
| Claude Opus 5 | $9.00 | 10% cheaper |
| Kimi K3 ouverte | $5.40 | 46% cheaper |
| Claude Opus 4.8 | $9.00 | 10% cheaper |
Questions fréquemment posées
How much does GPT-5.5 cost per 1M tokens?
GPT-5.5 costs $5.00 per 1M input tokens and $30.00 per 1M output tokens. At a typical 4:1 input-to-output mix that blends to about $10.00 per 1M tokens.
How much does GPT-5.5 cost per month?
A small-team workload of 20M input and 5M output tokens a month costs about $250 on GPT-5.5. A side project (1M in / 0.25M out) costs roughly $12.50.
What is a cheaper alternative to GPT-5.5?
Claude Opus 5 is the strongest cheaper option in our database at $9.00 per 1M blended — about 10% less than GPT-5.5.
Can I run GPT-5.5 locally?
No. GPT-5.5 is a closed, API-only model — the weights are not released, so it cannot be self-hosted.
Why does GPT-5.5 charge more for output than input?
Output tokens are generated one at a time and cannot be batched the way a prompt can, so they cost the provider more to serve. GPT-5.5 charges 6× more for output, which is why prompt-heavy workloads are far cheaper to run than generation-heavy ones.
Prices are the published list rates for the model's primary API and are reviewed as providers change them. Volume, batch and cached-input discounts are not included. Compare every model side by side in the Base de données des modèles d'IA ou le Classement des grands modèles linguistiques (LLM).
