GPT-6.1 Sol
What it is
OpenAI's GPT-6.1 Sol, positioned to balance intelligence and cost. It has a very large context window and cached-input pricing that helps with repeated system prompts.
Best for
#1 of 3Model for Build a deep-research agent
Total 7.5 / 10
- Task quality (40%)9 / 1095.4% on BrowseComp (Mercor 130); 67.7% on held-out Extended.
- Cost per typical job run (20%)8 / 10$2/$10 per MTok; $0.72 per AA index task.
Evidence:0082 (source: Artificial Analysis)0018 (source: OpenAI)
- Latency / speed (5%)5 / 1066.0 output tokens/s.
Evidence:0082 (source: Artificial Analysis)
- Tool-calling & structured-output reliability (15%)6 / 1070.0% on τ²-Airline.
Evidence:0014 (source: OpenRouter)
- Context & modality support (10%)8 / 101.05M context.
Evidence:0019 (source: OpenAI)
- Openness / deployability (10%)3 / 10Closed weights; OpenAI API.
Evidence:0018 (source: OpenAI)
#3 of 3Model for Build a customer-support agent
Total 6.0 / 10
- Task quality (15%)6 / 1070.0% τ²-Airline.
- Cost per typical job run (25%)7 / 10$0.10/task; $2/$10 per MTok.
- Latency / speed (15%)5 / 102.4 min/task.
Evidence:0014 (source: OpenRouter)
- Tool-calling & structured-output reliability (25%)6 / 10
Evidence:0014 (source: OpenRouter)
- Context & modality support (10%)8 / 101.05M context.
Evidence:0019 (source: OpenAI)
- Openness / deployability (10%)3 / 10Closed weights; OpenAI API.
Evidence:0018 (source: OpenAI)
#3 of 3Model for Ship a SaaS MVP in a weekend
Total 6.8 / 10
- Task quality (30%)8 / 10WebDev Arena 1755 (#4).
Evidence:0078 (source: LMArena)
- Cost per typical job run (20%)8 / 10$2/$10 per MTok; $0.72 per AA index task.
Evidence:0082 (source: Artificial Analysis)0018 (source: OpenAI)
- Latency / speed (15%)5 / 1066.0 output tokens/s.
Evidence:0082 (source: Artificial Analysis)
- Tool-calling & structured-output reliability (15%)6 / 1070.0% on τ²-Airline.
Evidence:0014 (source: OpenRouter)
- Context & modality support (10%)8 / 101.05M context.
Evidence:0019 (source: OpenAI)
- Openness / deployability (10%)3 / 10Closed weights; OpenAI API.
Evidence:0018 (source: OpenAI)
Strengths
- 1.05M-token context and 128K max output.0019 (source: OpenAI)
- Cached input at $0.10/MTok vs $2.00 uncached.0018 (source: OpenAI)
- Positioned to balance intelligence and cost.0019 (source: OpenAI)
Limits
- Proprietary, API-only model.0018 (source: OpenAI)
- Same headline price as Sonnet 5.5 ($2/$10 per MTok, short context), far above Haiku.0018 (source: OpenAI)
Pricing
Paid
$2.00 input / $0.10 cached / $10.00 output per MTok (short context).0018 (source: OpenAI)
As of . Check the vendor's page before buying.
License
Proprietary model (hosted access only).0156 (source: OpenAI)0157 (source: OpenAI)
Checked .
Alternatives in the same layer
Disclosures
- None