What other models does OpenAI offer?

OpenAI ships GPT-5.6 Sol alongside a broader lineup spanning multiple reasoning tiers, a prior-generation flagship, and an open-weight line for self-hosting. Here are some other models OpenAI supports:

• GPT-5.6 Terra: The balanced everyday tier of the 5.6 family, priced at roughly half of Sol's rate and tuned for standard coding, writing, and agentic tasks that don't need maximum reasoning effort. Runs at competitive quality to the prior GPT-5.5 flagship at a fraction of the cost

• GPT-5.6 Luna: The fastest and least expensive tier in the family, built for high-volume workloads like classification, summarization, and chat where throughput and price matter more than peak reasoning depth. Generates output at roughly triple Sol's tokens-per-second rate

• GPT-5.5: The prior-generation flagship, still available for teams that haven't migrated to 5.6. Shares the same 1M token context ceiling as Sol but carries an earlier knowledge cutoff and a lower coding-agent benchmark score

• GPT-5.4 mini: A smaller, lower-cost model from the 5.4 generation aimed at latency-sensitive product features rather than frontier reasoning. Fits use cases where a full-size reasoning model would be overkill

How does GPT-5.6 Sol differ from OpenAI's other models?

Within the GPT-5.6 family, Sol is the flagship reasoning tier, sitting above Terra and Luna on capability and price.

• Pricing: Sol costs $5 per million input tokens and $30 per million output tokens, versus $2.50 / $15 for Terra and $1 / $6 for Luna, so Terra runs at about half of Sol's rate and Luna at about a fifth

• Benchmark scores: Sol leads the family on the Artificial Analysis Intelligence Index at 59, ahead of Terra's 55 and Luna's 51, and separately leads the Artificial Analysis Coding Agent Index across all evaluated models in OpenAI's Codex harness

• Speed: Sol trades throughput for depth, generating output at roughly 69 tokens per second versus about 145 for Terra and nearly 200 for Luna, so the lighter tiers respond faster on latency-sensitive product surfaces

• Design and frontend work: Sol topped Design Arena's frontend leaderboard with an Elo of 1353, ahead of Claude Fable 5, a distinction OpenAI hasn't published for Terra or Luna

Sol fits workloads where reasoning depth and coding accuracy matter more than response speed or per-token cost: complex agentic pipelines, long-horizon coding tasks, and design or reasoning work where Terra and Luna's lighter compute budgets fall short.

What models should I consider using alongside GPT-5.6 Sol?

No single model is optimal for every task. Here are models worth pairing with GPT-5.6 Sol depending on what your product needs:

• Claude Opus 4.8: Reach for it on complex multi-step agentic reasoning where a second frontier-tier model can cross-check Sol's output or take over tasks where Anthropic's reasoning style scores better on your own evals

• Gemini 3.1 Pro Preview: Bring it in for long-context retrieval over very large document sets, where finding the right needle in a huge haystack matters more than reasoning from scratch

• Mistral Medium 3.5: Route to it when data residency or EU hosting requirements rule out a US-only provider, keeping sensitive workloads compliant without giving up much capability

• Grok 4.3: Pull it in for tasks that benefit from more current, real-time information grounding, useful when Sol's February 2026 knowledge cutoff is too stale for the query

What are the challenges of using GPT-5.6 Sol in my product?

Like any production LLM, GPT-5.6 Sol comes with tradeoffs worth planning for:

• Provider dependency: Routing every request straight to OpenAI means an outage, rate-limit change, or deprecation notice on Sol can take your product down with it unless you've built in a fallback path

• Cost at scale: At $5 / $30 per million tokens, Sol's reasoning-heavy usage compounds fast as volume grows, and teams that don't actively monitor spend can see bills climb well past what a lighter tier would have cost

• Latency: Sol's extended reasoning comes with a noticeably higher time to first token than Terra or Luna, which can hurt user-facing features that need a fast initial response

• Text-only output: Sol accepts image input but only generates text, so products that need native image or audio generation still need a second model in the pipeline

• Knowledge cutoff: Sol's training data ends in February 2026, so anything referencing more recent events or releases needs retrieval augmentation rather than relying on the model's own knowledge

Why should I use Merge Gateway to route LLM requests with GPT-5.6 Sol and every other model?

Using GPT-5.6 Sol through Merge Gateway gives you access to the model itself and the infrastructure layer around it:

• One API, every provider: Reach GPT-5.6 Sol and every other major LLM through a single endpoint and API key. Swap providers by changing the model string, with no application code changes required

• Intelligent routing and automatic failover: Merge routes around OpenAI outages automatically. Cost, latency, or quality-based routing policies can cut spend by 40-60% without touching your application code

• Cost governance: Set hard or soft project budgets so GPT-5.6 Sol spend stays within plan. Every request is attributed to a model, project, and tag in a unified billing dashboard across all providers

• Build Your Own Router: Decide what "best" means for your traffic by choosing from curated ML benchmarks or feeding in your own eval scores. The router scores each available model against your weights and picks a winner per request, with a plain-language explanation of every decision

• Security and compliance controls: Apply DLP rules and prompt injection protection before every request reaches OpenAI. Enforce per-project model and region policies without adding that logic to your application

How can I start routing requests to GPT-5.6 Sol via Merge Gateway?

Getting GPT-5.6 Sol running through Merge Gateway takes a few minutes:

1. Create an account and get your API key from the dashboard.

2. Install the Merge Gateway SDK: run pip install merge-gateway-sdk (Python) or npm install merge-gateway-sdk (Node). Alternatively, if you're already using the OpenAI SDK, set base_url = "https://api-gateway.merge.dev/v1/openai" and your existing code works as-is.

3. Make your first request using the provider/model format. For GPT-5.6 Sol, the model string is openai/gpt-5.6-sol. Swap the model string to route to any other provider without changing anything else.

4. Configure a routing policy in the dashboard to set failover behavior, cost limits, and optimization strategy. Your first policy can be as simple as naming GPT-5.6 Sol as primary with one fallback.

Full setup instructions and SDK references are in the Merge Gateway docs.