OpenAI has 6 models in The AI Leaderboard directory: GPT 5.6 Sol, GPT 5.5, GPT-5.4, GPT-5.3-Codex, GPT-5.2, gpt-oss-120b. This provider hub compares their verified availability, licensing, context windows, pricing references and intended uses, with direct links to every model guide.
Quick answer: GPT 5.6 Sol is OpenAI's highest-ranked model here at #3 with an Overall score of 81.0. Use the lineup table below to compare it with the provider’s other current, specialist, open-weight or migration-reference models.
OpenAI model lineup at a glance
Metric | Current directory value |
|---|---|
Models covered | 6 |
Currently ranked | 2 |
Open-weight models | 1 |
Restricted models | 0 |
Legacy or migration references | 0 |
Highest-ranked model | GPT 5.6 Sol, #3, Overall 81.0 |
Last verified | September 2, 2026 |
Complete OpenAI model directory
Model names in this table link directly to their complete guides. Status and pricing can change, so confirm the active endpoint and provider terms before procurement or migration.
Model | Status | License | Context | Best fit |
|---|---|---|---|---|
Available | Proprietary | 1.05M tokens | Reasoning, tool-using agents, coding and large repositories | |
Available | Proprietary | 1.05M tokens | Professional work, coding, analysis and adjustable reasoning | |
Available; previous flagship | Proprietary | 400,000 tokens | professional work, reasoning, coding, vision and long-context analysis | |
Available | Proprietary | 400,000 tokens | long-horizon agentic coding, repository work and technical professional tasks | |
Available | Proprietary | 1,050,000 tokens | professional work, computer-use agents, coding and large-context analysis | |
Available; downloadable weights | Open weights under Apache 2.0 | 131,072 tokens | self-hosted reasoning, tool use and production workloads that fit one 80GB GPU |
How the OpenAI models differ
Model | Directory role | Modalities | Maximum output | Access |
|---|---|---|---|---|
GPT 5.6 Sol | Available proprietary | Text and image input; text output | 128K tokens | See the guide for current access routes |
GPT 5.5 | Available proprietary | Text and image input; text output | 128K tokens | See the guide for current access routes |
GPT-5.2 | Available proprietary | Text and image input; text output | 128,000 tokens | ChatGPT and OpenAI API |
GPT-5.3-Codex | Available proprietary | Text and image input; text output | 128,000 tokens | Codex app, CLI, IDE extensions, web and OpenAI API |
GPT-5.4 | Available proprietary | Text and image input; text output | 128,000 tokens | ChatGPT as GPT-5.4 Thinking, OpenAI API and Codex |
gpt-oss-120b | Open weights | Text input; text output | Shares the model context budget | Hugging Face, compatible inference stacks and third-party hosts |
The labels above describe access and lifecycle, not a hidden capability ranking. Only 2 OpenAI models in the current top eight receive a rank on this site. Other guides remain unranked until comparable evidence is available.
OpenAI model pricing
Model | Pricing reference | Cost caveat |
|---|---|---|
GPT 5.6 Sol | $4 input / $20 output per million tokens, promotional | Caching, tools, long context and regional rates may differ |
GPT 5.5 | $5 input / $30 output per million tokens | Caching, tools, long context and regional rates may differ |
GPT-5.2 | $1.75 input, $0.175 cached input and $14 output per million tokens | Caching, tools, long context and regional rates may differ |
GPT-5.3-Codex | Usage follows current Codex and API pricing; check the OpenAI model card for the active rate | Caching, tools, long context and regional rates may differ |
GPT-5.4 | $2.50 input, $0.25 cached input and $15 output per million tokens; long-context multipliers apply above 272K input tokens | Caching, tools, long context and regional rates may differ |
gpt-oss-120b | No OpenAI token fee for self-hosting; infrastructure and hosted-provider rates vary | Self-hosting still incurs infrastructure and operations costs |
Token price is not cost per accepted result. Measure retries, tool calls, latency, infrastructure and reviewer time on the same task set.
Which OpenAI model should you choose?
Requirement | Starting point | Why |
|---|---|---|
Highest current rank | GPT 5.6 Sol | Highest OpenAI placement in the current top eight |
Open weights or self-hosting | gpt-oss-120b | Review the exact license, checkpoint and hardware needs |
Restricted specialist work | No restricted model in this provider directory | Use generally available models and least-privilege controls |
Migration or reproduction | gpt-oss-120b | Compare current versions using pinned snapshots |
Lowest operational cost | Measure two or three eligible models | Headline token prices omit retries, tools, caching and human review |
OpenAI models, explained
GPT 5.6 Sol
GPT 5.6 Sol was released July 9, 2026. It is listed as available with proprietary. The documented context window is 1.05M tokens, and the maximum output is 128K tokens.
- Best fit: Reasoning, tool-using agents, coding and large repositories.
- Modalities: Text and image input; text output.
- Access: See the guide for current access routes.
- Pricing reference: $4 input / $20 output per million tokens, promotional.
- Current ranking: #3 with an Overall score of 81.0.
Read the complete GPT 5.6 Sol guide for benchmarks, limitations, official sources and an evaluation framework.
GPT 5.5
GPT 5.5 was released April 24, 2026 (API). It is listed as available with proprietary. The documented context window is 1.05M tokens, and the maximum output is 128K tokens.
- Best fit: Professional work, coding, analysis and adjustable reasoning.
- Modalities: Text and image input; text output.
- Access: See the guide for current access routes.
- Pricing reference: $5 input / $30 output per million tokens.
- Current ranking: #4 with an Overall score of 80.2.
Read the complete GPT 5.5 guide for benchmarks, limitations, official sources and an evaluation framework.
GPT-5.2
GPT-5.2 was released December 11, 2025. It is listed as available; previous flagship with proprietary. The documented context window is 400,000 tokens, and the maximum output is 128,000 tokens.
- Best fit: professional work, reasoning, coding, vision and long-context analysis.
- Modalities: Text and image input; text output.
- Access: ChatGPT and OpenAI API.
- Pricing reference: $1.75 input, $0.175 cached input and $14 output per million tokens.
Read the complete GPT-5.2 guide for benchmarks, limitations, official sources and an evaluation framework.
GPT-5.3-Codex
GPT-5.3-Codex was released February 5, 2026. It is listed as available with proprietary. The documented context window is 400,000 tokens, and the maximum output is 128,000 tokens.
- Best fit: long-horizon agentic coding, repository work and technical professional tasks.
- Modalities: Text and image input; text output.
- Access: Codex app, CLI, IDE extensions, web and OpenAI API.
- Pricing reference: Usage follows current Codex and API pricing; check the OpenAI model card for the active rate.
Read the complete GPT-5.3-Codex guide for benchmarks, limitations, official sources and an evaluation framework.
GPT-5.4
GPT-5.4 was released March 5, 2026. It is listed as available with proprietary. The documented context window is 1,050,000 tokens, and the maximum output is 128,000 tokens.
- Best fit: professional work, computer-use agents, coding and large-context analysis.
- Modalities: Text and image input; text output.
- Access: ChatGPT as GPT-5.4 Thinking, OpenAI API and Codex.
- Pricing reference: $2.50 input, $0.25 cached input and $15 output per million tokens; long-context multipliers apply above 272K input tokens.
Read the complete GPT-5.4 guide for benchmarks, limitations, official sources and an evaluation framework.
gpt-oss-120b
gpt-oss-120b was released August 5, 2025. It is listed as available; downloadable weights with open weights under apache 2.0. The documented context window is 131,072 tokens, and the maximum output is Shares the model context budget.
- Best fit: self-hosted reasoning, tool use and production workloads that fit one 80GB GPU.
- Modalities: Text input; text output.
- Access: Hugging Face, compatible inference stacks and third-party hosts.
- Pricing reference: No OpenAI token fee for self-hosting; infrastructure and hosted-provider rates vary.
Read the complete gpt-oss-120b guide for benchmarks, limitations, official sources and an evaluation framework.
How to evaluate OpenAI models
Shortlist the OpenAI models that meet your access, licensing and modality requirements. Freeze 20 to 50 representative tasks, then give each candidate the same prompts, tools, context, reasoning budget and retry limits. Record the exact model identifier because provider aliases can change.
Measure | Record | Decision use |
|---|---|---|
Task completion | Pass rate and rubric score | Establish whether the model meets the acceptance threshold |
Reliability | Repeated-run variance and recovery behavior | Avoid choosing from one lucky result |
Quality | Factuality, instruction following and reviewer acceptance | Separate polished output from correct output |
Tool use | Wrong calls, retries, permission errors and unsafe actions | Evaluate the model and agent harness together |
Efficiency | Wall time, tokens, caching, tools and review labor | Calculate cost per accepted result |
Governance | Retention, training, regional processing and audit logs | Confirm the deployment fits policy requirements |
Availability, versions and lifecycle
OpenAI may expose different snapshots, effort settings, products and partner integrations under similar labels. Treat every model name as incomplete without an exact endpoint or checkpoint. Re-run regression tests after any model, system prompt, retrieval, tool or permission change.
- Available means a documented access route exists, not that every region or account has access.
- Preview models can change behavior, price or retirement dates before general availability.
- Open weights require an exact license and checkpoint review; “open” does not guarantee open training data or low operating cost.
- Restricted models require separate eligibility, safeguards and data-handling review.
- Legacy models remain useful for migration and reproduction but should not be the default for a new deployment.
Sources and update policy
Official OpenAI model documentation is the provider-level starting point. Each linked model guide cites its own first-party launch page, model card, pricing reference and safety documentation.
Compare this provider with the complete Best AI Models directory.
Last verified September 2, 2026. The hub should be updated when the provider launches, retires, renames or reprices a model, changes context or output limits, alters licensing, or modifies regional availability.
Frequently asked questions
How many OpenAI models are covered?
This hub covers 6: GPT 5.6 Sol, GPT 5.5, GPT-5.4, GPT-5.3-Codex, GPT-5.2, gpt-oss-120b. Every model has a dedicated guide linked from the directory table and profile sections.
What is the best OpenAI model?
GPT 5.6 Sol is OpenAI's highest-ranked model in the current AI Leaderboard, at number 3 with an Overall score of 81.0. Workload-specific testing can still favor another model.
Does OpenAI offer open weights?
Yes. The directory identifies gpt-oss-120b as open-weight options. Review the exact license, downloadable artifact and serving requirements before deployment.
How current is the pricing?
Pricing references were verified September 2, 2026 from first-party documentation summarized in each model guide. Always recheck the provider before budgeting because cache, batch, long-context, tool and regional rates can change.
Should every OpenAI model be tested?
No. Remove models that fail access, license, modality, context, latency or governance requirements first. Test the smallest credible shortlist on representative work.