OpenAI
GPT series, o-series reasoning models, and multimodal APIs.
StatusOperational Live status
LLMAPI discountUp to −25%
Models on LLMAPI25
From$0.14 / 1M blended
RegionsUS, EU
Best uptime (24h)100.00%
Verdict
OpenAI develops the GPT and o-series model families, the most widely adopted LLM APIs in production, spanning chat, reasoning, vision, audio, and embeddings. Through LLMAPI you reach 15 OpenAI models on the same OpenAI-compatible endpoint as every other provider, so switching between them is a one-string change.
Good fit for
- Text Generation
- Code Generation
- Vision / Image Understanding
- Speech-to-Text
- Text-to-Speech / Voice
- Multimodal (any-to-any)
Consider another provider for
- No image generation models in this catalog
- No video generation models in this catalog
- No video understanding models in this catalog
- No music & audio generation models in this catalog
Key facts
Where a provider does not document something we say so rather than guess. Everything LLMAPI records about OpenAI, in one table.
Overview | |
| Provider | OpenAI |
| Website | openai.com |
| Best for | General-purpose chat, multimodal apps, and the widest tooling ecosystem |
| Models on LLMAPI | 25 |
| Popular models | GPT-6 Astra, GPT Live 1, GPT-5.6 Luna, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.5 |
Access & routing | |
| API compatibility | OpenAI-compatible — /v1/chat/completions |
| Model ID prefix | openai/<model> |
| Regions | US, EU |
| Fallback | Configurable — on 429 or 5xx the router switches to any other model or provider on your list |
| Billing | One invoice across every provider, with per-key spend limits |
| Volume discount | Eligible |
Performance (median across listed models) | |
| Blended price / 1M | $3.5 |
| Uptime (24h) | 99.94% (median across listed models) |
Modality | |
| Tags | |
Capability | |
| Tags | |
Workflow | |
| Tags | |
Audience | |
| Tags | |
Other | |
| Tags | |
Privacy & data | |
| Inference location | US and EU regions, selected per endpoint |
| Data residency | EU data residency available on eligible plans |
| Default retention | 30 days for abuse monitoring, then deleted |
| Zero-data retention | On request Zero data retention for eligible endpoints and accounts |
| Training on API data | No API data is not used to train models by default |
| Human review | Flagged abuse cases only |
Compliance & lifecycle | |
| Certifications | SOC 2 Type II · ISO 27001 · GDPR DPA · HIPAA BAA available |
| Deprecation notice | Announced in the deprecations changelog |
| Operational status | Operational Live status |
Prices and context from the OpenRouter public catalogue; uptime and output speed from OpenRouter provider endpoints; quality scores from LiveBench. Refreshed nightly — last updated 2026-09-29. — means no figure is published for that model.
Best OpenAI models
A quick view of how OpenAI 's 15 models compare on intelligence, output speed and price — so you can pick the right one for your use case.
Most intelligent
#1GPT-4o TTS40
#2GPT-4o39
#3GPT-5 mini37
#4text-embedding-3-large30
Intelligence index · 15 models
Fastest
#1o4-mini441 t/s
#2text-embedding-3-small428 t/s
#3Whisper424 t/s
#4GPT-4.1386 t/s
Output tokens / second · 15 models
Lowest price
#1text-embedding-3-small$1.00
#2o3$1.72
#3GPT-4.1$2.50
#4o4$3.19
Blended price per 1M tokens · 15 models
Compared with other hosts
How OpenAI sits against the rest of the catalog, using the same measurements shown on every provider page.
- Price. Median blended price of $7.75 per 1M tokens — 26% above the $6.13 median across all 50 providers in the catalog.
- Throughput. Median output speed of 348 tokens/s, 0% slower than the 348 tok/s catalog median.
- Catalog depth. 15 models listed, against a catalog average of 5 per provider.
- Integration. Identical to every other provider here — same endpoint, same SDK, same keys, so a switch costs one string change.
Median blended price per 1M tokens
lower is better| OpenAI | $7.75 | |
| Microsoft Azure | $7.75 | |
| Together AI | $7.80 | |
| Liquid AI | $7.61 | |
| Replicate | $8.16 | |
| Inco | $7.33 |
Median output speed
higher is better| OpenAI | 348 tok/s | |
| Microsoft Azure | 202 tok/s | |
| Together AI | 781 tok/s | |
| Liquid AI | 233 tok/s | |
| Replicate | 748 tok/s | |
| Inco | 697 tok/s |
Figures are medians across the models listed on each provider page and are refreshed with the catalog.
Models & pricing
Every OpenAI model reachable with your LLMAPI key. Click a column heading to sort.
| Model | API model name | Uptime 24h | Price / 1M | Context |
|---|---|---|---|---|
| GPT-6 Astra | gpt-6-astra | 100.00% | $ 20 | 1050 k |
| GPT Live 1 | gpt-live-1 | — | $ 0 | — |
| GPT-5.6 Luna | gpt-5.6-luna | 99.99% | $ 0.45 | 1050 k |
| GPT-5.6 Sol | gpt-5.6-sol | 99.94% | $ 8 | 1050 k |
| GPT-5.6 Terra | gpt-5.6-terra | 99.97% | $ 4.5 | 1050 k |
| GPT-5.5 | gpt-5.5 | 100.00% | $ 11.25 | 1050 k |
| GPT-5.4 Nano | gpt-5.4-nano | 99.94% | $ 0.46 | 400 k |
| GPT-5.4 Mini | gpt-5.4-mini | 99.95% | $ 1.69 | 400 k |
| GPT-5.4 | gpt-5.4 | 100.00% | $ 5.63 | 1050 k |
| GPT-5.4 Pro | gpt-5.4-pro | 99.84% | $ 67.5 | 1050 k |
| GPT-5.3 Codex | gpt-5.3-codex | 99.98% | $ 4.81 | 400 k |
| GPT-5.2 | gpt-5.2 | 99.92% | $ 4.81 | 400 k |
| GPT-5.2 Pro | gpt-5.2-pro | 99.94% | $ 57.75 | 400 k |
| GPT-5.1 | gpt-5.1 | 99.89% | $ 3.44 | 400 k |
| GPT-5 | gpt-5 | 99.90% | $ 3.44 | 400 k |
| GPT-5 Mini | gpt-5-mini | 99.92% | $ 0.69 | 400 k |
| GPT-5 Pro | gpt-5-pro | 100.00% | $ 41.25 | 400 k |
| GPT-5 Nano | gpt-5-nano | 99.93% | $ 0.14 | 400 k |
| o3 | o3 | 100.00% | $ 3.5 | 200 k |
| GPT-4.1 | gpt-4.1 | 99.90% | $ 3.5 | 1000 k |
| GPT-4.1 Nano | gpt-4.1-nano | 99.89% | $ 0.18 | 1000 k |
| GPT-4.1 Mini | gpt-4.1-mini | 99.84% | $ 0.7 | 1000 k |
| GPT-4o Mini Transcribe | gpt-4o-mini-transcribe | — | $ 0 | — |
| GPT-4o Mini Search Preview | gpt-4o-mini-search-preview | — | $ 0.26 | 128 k |
| GPT-4o Transcribe | gpt-4o-transcribe | — | $ 0 | — |
Price is a blended per-1M-token figure; latency is time to first token and throughput is output tokens per second, measured on the LLMAPI edge.
Cost calculator
Pick a model, enter your traffic, and see the monthly bill at LLMAPI rates.
Estimates use the blended per-1M-token price shown in the table above. Word counts differ by language — Cyrillic and CJK text uses 2–3× more tokens per word.
OpenAI via LLMAPI
Same models, same features, one key. Prefix the model ID with openai/ and point your OpenAI client at our base URL.
# Python · OpenAI SDK from openai import OpenAI client = OpenAI( base_url="https://api.llmapi.ai/v1", api_key="LLMAPI_KEY", ) r = client.chat.completions.create( model="openai/gpt-5", messages=[{"role": "user", "content": "Hello"}], )
| Pricing | Eligible for LLMAPI volume discounts; billed on one invoice with every other provider. |
| Features | Full pass-through — tool calling, JSON schema, streaming, vision and reasoning behave exactly as on OpenAI’s own API. |
| Fallback | Configurable. On 429 or 5xx the router switches to any model or provider on your fallback list. |
| Regions | US, EU |
| Your data | LLMAPI stores request metadata only — model, token counts, latency and status. No prompt or completion content. |
Capabilities & use cases
Each tag below is its own catalog page listing every provider that shares it.
Text GenerationCode GenerationVision / Image UnderstandingSpeech-to-TextText-to-Speech / VoiceMultimodal (any-to-any)Instruction FollowingFunction Calling / Tool UseAgentic Task ExecutionCreative WritingCoding AutocompleteAutonomous Software EngineeringDebugging & Code ReviewRAG & Knowledge RetrievalWeb Research AgentsEmail DraftingBrainstorming & IdeationBest for BeginnersBest for DevelopersMost PopularBest Model for Building Agents
Developer resources
Where to go next.
Join thousands of developers building on one API key
Every provider in the catalog, one key, one invoice.
Questions
How do I call OpenAI models through LLMAPI?
Point any OpenAI-compatible client at https://api.llmapi.ai/v1 and use the model ID openai/gpt-5. No other change is needed.
Does OpenAI cost more through LLMAPI?
No. This provider is eligible for LLMAPI volume discounts, so heavy usage lands below list price.
Which regions are used?
US, EU
What happens if the provider returns an error?
On 429 or 5xx responses the router falls back to any other model or provider on your configured list, so requests keep succeeding.
How many OpenAI models are available?
15 at the moment, all listed in the models table above and updated as the provider ships new ones.
Related providers
Providers covering similar use cases.