GPT-6.2 Sol vs Mercury 2.5 Preview
OpenAI · US | Inception Labs · US · Updated November 2026
Quick verdict
Mercury 2.5 Preview is the value pick at $0.04/$0.15 per 1M tokens, about 50× cheaper on input than GPT-6.2 Sol ($2/$10 per 1M tokens). For very long inputs, GPT-6.2 Sol's 1.05M tokens window beats Mercury 2.5 Preview's 260K tokens. GPT-6.2 Sol is the newer release (November 17, 2026).
GPT-6.2 Sol (OpenAI) and Mercury 2.5 Preview (Inception Labs) are two of the models people most often weigh against each other in 2026. GPT-6.2 Sol is OpenAI's November 17, 2026 efficient-tier refresh - 1.05M context at $2/$10 per million tokens. Mercury 2.5 Preview is Inception Labs' August 31, 2026 diffusion-based language model preview, refining tokens in parallel for roughly 10x the throughput of comparable autoregressive models at cost-optimized-tier quality. They diverge most on price and context window — each quantified below from the models' real specs.
Key differences at a glance
- ▸Price: Mercury 2.5 Preview is about 50× cheaper on input ($0.04/$0.15 per 1M tokens vs $2/$10 per 1M tokens) — a large enough gap that at scale it can be the single biggest line item in the decision.
- ▸Context window: GPT-6.2 Sol holds 4× more — 1.05M tokens (~1,575 pages) vs 260K tokens (~390 pages). But effective recall usually fades long before the advertised ceiling, so the bigger number only helps if the model reasons over it.
- ▸Recency: GPT-6.2 Sol is the newer model by about 3 months (released November 17, 2026), usually meaning fresher training data and capabilities.
Side-by-side specs
| Spec | GPT-6.2 Sol | Mercury 2.5 Preview |
|---|---|---|
| Provider | OpenAI (US) | Inception Labs (US) |
| Released | November 17, 2026 | August 31, 2026 |
| Context window | 1.05M tokens (~1,575 pages) | 260K tokens (~390 pages) |
| Price (in/out) | $2/$10 per 1M tokens | $0.04/$0.15 per 1M tokens |
| Open weight? | No — API only | No — API only |
| Modalities | text, image | text |
| SWE-Bench Verified | Not published | Not published |
| MRCR v2 @ 1M | Not published | Not published |
Who wins what
OpenAI's efficient everyday tier at $2/$10 per million tokens, unchanged from GPT-6.1 Sol
GPT-6.2 Sol
Its 1.05M tokens window holds about 4× more than Mercury 2.5 Preview's 260K tokens in a single prompt.
Artificial Analysis Intelligence Index of 57.3 (as cited by TestingCatalog)
GPT-6.2 Sol
OpenAI's November 17, 2026 efficient-tier refresh - 1.05M context at $2/$10 per million tokens — and it carries the larger 1.05M tokens context.
1.05M-token context with 922K max input and 128K max output; cached input $0.10/MTok
GPT-6.2 Sol
Mercury 2.5 Preview is comparatively weak here — 260K context window is far shorter than frontier 1M-token models
Very high inference throughput (1,107 tok/s) via diffusion-based parallel token generation
Mercury 2.5 Preview
Inception Labs' August 31, 2026 diffusion-based language model preview, refining tokens in parallel for roughly 10x the throughput of comparable autoregressive models at cost-optimized-tier quality — and it runs cheaper at $0.04/$0.15 per 1M tokens.
Coding accuracy (95.7%, 91st percentile among cost-optimized models)
Mercury 2.5 Preview
At $0.04/$0.15 per 1M tokens it undercuts GPT-6.2 Sol ($2/$10 per 1M tokens), and that gap compounds at volume.
Mathematics accuracy (97.0%, 97th percentile)
Mercury 2.5 Preview
Mercury 2.5 Preview lists mathematics accuracy (97.0%, 97th percentile) among its strengths; GPT-6.2 Sol does not.
Lowest cost at scale
Mercury 2.5 Preview
At $0.04/$0.15 per 1M tokens, it is the cheaper of the two — the gap dominates the bill on high-volume workloads.
Largest single-prompt input
GPT-6.2 Sol
Its 1.05M tokens window is about 4× larger than Mercury 2.5 Preview's 260K tokens, fitting roughly 1,575 pages in one prompt.
Which should you pick?
A cost-sensitive startup shipping high volume
→ Mercury 2.5 Preview
At $0.04/$0.15 per 1M tokens it undercuts GPT-6.2 Sol, and on millions of tokens that margin decides the monthly bill.
Someone analysing very long documents or codebases
→ GPT-6.2 Sol
Its 1.05M tokens window holds about 4× more than Mercury 2.5 Preview's 260K tokens in one prompt.
Anyone who needs image input
→ GPT-6.2 Sol
GPT-6.2 Sol lists image among its modalities; Mercury 2.5 Preview does not.
Someone who wants the most recent release
→ GPT-6.2 Sol
Released November 17, 2026, after Mercury 2.5 Preview (August 31, 2026).
If this matters most: OpenAI's efficient everyday tier at $2/$10 per million tokens, unchanged from GPT-6.1 Sol
→ GPT-6.2 Sol
This is GPT-6.2 Sol's main documented strength.
If this matters most: Very high inference throughput (1,107 tok/s) via diffusion-based parallel token generation
→ Mercury 2.5 Preview
This is Mercury 2.5 Preview's main documented strength.
GPT-6.2 Sol: where it fits
OpenAI's November 17, 2026 efficient-tier refresh - 1.05M context at $2/$10 per million tokens. Released November 17, 2026 by OpenAI.
Strengths: OpenAI's efficient everyday tier at $2/$10 per million tokens, unchanged from GPT-6.1 Sol; Artificial Analysis Intelligence Index of 57.3 (as cited by TestingCatalog); 1.05M-token context with 922K max input and 128K max output; cached input $0.10/MTok; Available in ChatGPT Work, Codex and the API as gpt-6.2-sol.
Trade-offs: Below GPT-6.2 Astra on reasoning-heavy work; Text and image input only; no audio or video, and text-only output; Same list price as GPT-6.1 Sol, so the update is a capability gain rather than a price cut. At $2 in / $10 out per million tokens, it sits in the mid price band.
Mercury 2.5 Preview: where it fits
Inception Labs' August 31, 2026 diffusion-based language model preview, refining tokens in parallel for roughly 10x the throughput of comparable autoregressive models at cost-optimized-tier quality. Released August 31, 2026 by Inception Labs.
Strengths: Very high inference throughput (1,107 tok/s) via diffusion-based parallel token generation; Coding accuracy (95.7%, 91st percentile among cost-optimized models); Mathematics accuracy (97.0%, 97th percentile); Tunable reasoning levels with parallel tool calls and schema-aligned JSON output.
Trade-offs: 260K context window is far shorter than frontier 1M-token models; No vision or audio modalities; Positioned only against cost-optimized models (GPT-5.6 Luna Low, Gemini 3.5 Flash-Lite), not frontier-class; Current $0.04/$0.15 pricing includes a limited-time launch promotion on OpenRouter (list price is $0.20/$0.75, may rise after Sept 8, 2026). At $0.04 in / $0.15 out per million tokens, it sits in the budget price band.
The bottom line for this matchup
GPT-6.2 Sol and Mercury 2.5 Preview overlap enough that the right pick depends on your specific job. Mercury 2.5 Preview costs less per token. GPT-6.2 Sol holds the larger context. Rather than crowning one, run the same hard task through both once and let the results decide.
Comparing models for real work? LumiChats puts 40+ leading AI models under one ₹69/day pass (about $1/day), so you can draft with one and cross-check with another without stacking subscriptions.
See pricingFrequently asked questions
Is GPT-6.2 Sol or Mercury 2.5 Preview better for coding?
Public SWE-Bench figures are not available for either model, so the honest test is your own repository — run an identical real bug through both.
Which is cheaper, GPT-6.2 Sol or Mercury 2.5 Preview?
Mercury 2.5 Preview is cheaper — $2/$10 per 1M tokens vs $0.04/$0.15 per 1M tokens, roughly 50× apart on input.
Which has the bigger context window?
GPT-6.2 Sol — 1.05M tokens vs 260K tokens, about 4× larger. Useful only if the model actually reasons over the full window, which not all do.
Can I use both GPT-6.2 Sol and Mercury 2.5 Preview together?
Yes. Many teams draft with one model and check the result with another. You can call each through its own provider, or use a multi-model platform; LumiChats, for example, puts 40+ leading models under one ₹69/day pass (about $1/day). Check which models a platform currently offers before you subscribe.
Which is newer, GPT-6.2 Sol or Mercury 2.5 Preview?
GPT-6.2 Sol — released November 17, 2026, about 3 months after Mercury 2.5 Preview.
Related comparisons
New model matchups
Sources
Specifications and benchmarks reflect publicly reported figures as of November 2026 and may change as providers release updates. Always verify on your own workload.