GPT-5.6 Luna vs Inkling-Small: Specs & Benchmark Comparison

CharacteristicGPT-5.6 LunaInkling-Small
CompanyOpenAIThinking Machines Lab
Release DateJuly 9, 2026July 30, 2026
Parameters276B
MultimodalYesYes
Context (input)1.1M256K
Context (output)128K256K
Input Price / 1M$0.20$0.30
Output Price / 1M$1.20$1.20
Average Score0.50.6
Benchmarks
Terminal-Bench 2.10.80.6
Artificial Analysis0.50.4
SWE-Bench Pro0.60.6
BrowseComp0.80.8
MMMU-Pro0.80.7
GPQA0.90.9
Toolathlon0.50.5

Visual Benchmark Comparison

GPT-5.6 Luna
Inkling-Small
Terminal-Bench 2.10.8 vs 0.6
0.8
0.6
Artificial Analysis0.5 vs 0.4
0.5
0.4
SWE-Bench Pro0.6 vs 0.6
0.6
0.6
BrowseComp0.8 vs 0.8
0.8
0.8
MMMU-Pro0.8 vs 0.7
0.8
0.7
GPQA0.9 vs 0.9
0.9
0.9
Toolathlon0.5 vs 0.5
0.5
0.5

Verdict

GPT-5.6 Luna leads in 2 out of 4 comparison categories.

Overall Performance

Both models show comparable average scores: GPT-5.6 Luna — 0.5, Inkling-Small — 0.6.

API Cost

GPT-5.6 Luna is 1.1x cheaper: input $0.20/1M vs $0.30/1M tokens.

Context Window

GPT-5.6 Luna supports a larger context: 1M vs 256K tokens.

Recency

Inkling-Small is newer: released 7/30/2026 vs 7/9/2026.

More About These Models

Related Comparisons

Frequently Asked Questions

Which is better for coding — GPT-5.6 Luna or Inkling-Small?
Direct comparison on the SWE-Bench benchmark is not available. We recommend reviewing other metrics on the comparison page.
Which model is cheaper — GPT-5.6 Luna or Inkling-Small?
GPT-5.6 Luna is cheaper for input: $0.20 per 1M tokens vs $0.30.
Which has a larger context window — GPT-5.6 Luna or Inkling-Small?
GPT-5.6 Luna supports a larger context: 1,050,000 tokens vs 256,000.

The GPT-5.6 Luna and Inkling-Small comparison is updated for 2026. Data includes benchmark results, API pricing, context window size and other specifications. For more detailed information, visit the GPT-5.6 Luna or Inkling-Small page. See also the complete list of AI model comparisons.