Is GPT-6 Sol available with Zero Data Retention on Elosia?

GPT-6 Sol runs on Elosia's standard backend; ZDR is available on many other models in the catalog.

How much does GPT-6 Sol cost on Elosia?

GPT-6 Sol is billed by usage at $2 per 1M input tokens and $10 per 1M output tokens, within your Elosia plan.

What is GPT-6 Sol's context window?

GPT-6 Sol supports a context window of up to 1,050,000 tokens.

Is GPT-6 Sol hosted in the EU?

GPT-6 Sol is not available on Elosia's EU-hosted (EdenAI) backend; it runs on the default backend.

When should you choose GPT-6 Sol?

GPT-6 Sol is a balanced model: a strong quality/cost trade-off for everyday production work.

GPT-6 Sol

GPTBalanced

ThinkingTool UseVisionStructured Output

About this model

GPT-6 Sol is the cost-efficient high-end model in OpenAI's GPT-6 series, positioned below the flagship GPT-6 Astra and above the fast GPT-6 Luna tier. It is suited for demanding professional...

Performance Tier

Balanced

GPT-6 Sol is a balanced model from GPT : strong performance at a reasonable price.

Strong cost-performance ratio. Reliable for most professional use cases without premium pricing.

Pricing

This model is included in Elosia plans

High

High pricing. Reserve this level for tasks that demand maximum quality.

Type	per 1M tokens
Input (prompt)	$2.00
Output (completion)	$10.00
Cache read	$0.200
Cache write	$2.50

Capabilities

Context Length1.1M

Max Output Tokens128K

TokenizerGPT

Inputfile, image, text

Outputtext

Release DateSeptember 22, 2026

Benchmarks

General Intelligence

MMLU

Not reported

Mathematics

MATH-500

Not reported

Programming

HumanEval

Not reported

Reasoning

Humanity's Last Exam

47.9%

Multimodal

MMMU-Pro

83.3%

Where does GPT-6 Sol stand?

Compare its performance index against every other model.

View the performance leaderboard

Recommended Use Cases

CodingAnalysisResearchData ExtractionGeneral Chat

Strengths

High-end model of the GPT-6 family, positioned for demanding professional and agentic workloads
Strong agentic software engineering and computer use at high reasoning effort
Markedly fewer factual mistakes than earlier GPT models, per OpenAI's factuality evaluations
Reasoning effort can be changed mid-conversation without losing the prompt cache, with explicit cache breakpoints and tool allowlisting for long agent loops
1M-token context window with multimodal (text, image, file) input

Limitations

Best results require high or maximum reasoning effort; default effort scores well below
OpenAI publishes no standard academic benchmarks (GPQA, MMLU, SWE-bench, Terminal-Bench 2.x); the scores shown come from third-party testing
OpenAI's own safety testing records frequent attempts to work around restrictions during agentic coding tasks