Moonshot AI open-weight challenger vs OpenAI flagship · research snapshot July 18, 2026

Kimi K3 vs GPT-5.6 Sol

Kimi K3 is the largest open-weight model ever at 2.8 trillion parameters. GPT-5.6 Sol is OpenAI's new flagship — the top tier of the GPT-5.6 family, with an Ultra mode that hits 91.9% on Terminal-Bench 2.1. Sol leads on 6 of 9 shared benchmarks; K3 leads on 3 and costs 40% less per token. But the real story is how close they are: this is the first time an open-weight model has traded blows with OpenAI's best. Here's the full comparison using Moonshot's launch table, OpenAI's system card, and independent benchmarks.

Last updated July 18, 2026 · K3 weights promised by July 27 Sol: GA since July 9, $5/$30 per 1M

88.3%
K3 Terminal-Bench 2.1 (KimiCode)
88.8%
Sol Terminal-Bench 2.1 (Codex)
$3 / $15
K3 input / output per 1M
$5 / $30
Sol input / output per 1M
Short answer: pick GPT-5.6 Sol for maximum coding capability, controlled reasoning, and the OpenAI ecosystem — it leads on 6 of 9 shared benchmarks including DeepSWE (+5.5) and GDPval-AA (+80 Elo). Pick Kimi K3 for frontend coding (#1 on Arena), long-horizon agents, cost-sensitive pipelines, and open-weight flexibility — it leads on BrowseComp, AutomationBench, and Toolathlon, and costs 40% less. The gap is narrow enough that workload routing beats picking a single winner.

Model cards side by side

SpecificationKimi K3GPT-5.6 Sol
Provider / releaseMoonshot AI · July 16, 2026OpenAI · July 9, 2026 GA
API model IDkimi-k3gpt-5.6-sol (alias: gpt-5.6)
ArchitectureMoE · 2.8T total / 16 of 896 experts activeProprietary (undisclosed)
Attention mechanismKimi Delta Attention + Attention ResidualsNot publicly detailed
Context window1,048,576 tokens (1M)1.05M tokens
Maximum output131K default, up to 1,048,576128K tokens
Reasoning modesMax effort only at launch; lower modes plannedNone through Max + Ultra mode (parallel multi-agent)
Input modalitiesText, image, videoText, image
ToolsFunction calling, structured outputsFunctions, web search, file search, computer use, MCP
Weights / licenseOpen weights (Modified MIT), promised by July 27Closed, proprietary, API-only
Input / output price$3 / $15 per 1M$5 / $30 per 1M
Cached input$0.30 per 1M, automatic$0.50 per 1M; writes at 1.25× input rate
Ultra / high-effort modeNot availableSol Ultra: 91.9% Terminal-Bench 2.1
Speed~62 tok/s (Artificial Analysis)~53 tok/s; up to 750 tok/s on Cerebras

Sol is the flagship of OpenAI's three-tier GPT-5.6 family (Sol, Terra, Luna). It ships with an Ultra mode that coordinates parallel sub-agents — the only model in this comparison with that capability. K3 counters with published architecture, planned open weights, native video input, and a 40% price advantage. Sol has the richer tool ecosystem (web search, file search, computer use); K3 has function calling and structured outputs.

Benchmarks: Sol leads on 6 of 9 shared, but K3 wins key agentic rows

The table below draws from Moonshot's official K3 launch comparison, OpenAI's GPT-5.6 system card, BenchLM shared results, and independent leaderboards. Harnesses differ per row — Moonshot uses KimiCode for K3 and Codex for Sol on several coding benchmarks. Treat small gaps as near-ties.

BenchmarkKimi K3GPT-5.6 SolLeader
Terminal-Bench 2.188.388.8Sol +0.5 · near-tie; different harnesses
DeepSWE67.573.0Sol +5.5 · repository engineering
FrontierSWE81.271.3K3 +9.9 · largest K3 lead on any chart
BrowseComp91.290.4K3 +0.8 · narrow; Sol Ultra: 92.2
GDPval-AA v2 (Elo)1,6681,748Sol +80 Elo · professional knowledge work
GPQA Diamond93.594.6Sol +1.1 · both near ceiling
AutomationBench30.829.1 (AA: 51.2)K3 +1.7 · different sources
Toolathlon-Verified73.2Not in Moonshot tableK3-only evidence
AA-Briefcase (Elo)1,5481,495K3 +53 Elo · independent
Key coding & agent benchmarks: K3 vs Sol
Higher is better. Sources: Moonshot K3 launch table, OpenAI GPT-5.6 system card, BenchLM, Artificial Analysis.
Terminal-Bench · K3
88.3
Terminal-Bench · Sol
88.8
DeepSWE · K3
67.5
DeepSWE · Sol
73.0
FrontierSWE · K3
81.2
FrontierSWE · Sol
71.3
Harness hygiene: Moonshot's table uses KimiCode for K3 and Codex for Sol on Terminal-Bench 2.1 and DeepSWE. FrontierSWE uses KimiCode for K3 and Codex for Sol. These are different agent scaffolds. Sol also has an Ultra mode (parallel multi-agent) that scores 91.9% on Terminal-Bench 2.1 — not shown in the base comparison. The independent Artificial Analysis Intelligence Index places Sol (max) at 58.9 vs K3 at ~57. Use these numbers to inform testing, not to declare a universal winner.

OpenAI's system card: what Sol brings that K3 doesn't

BenchmarkGPT-5.6 SolContext
Terminal-Bench 2.1 (Ultra)91.9%Parallel multi-agent; beats Mythos 5 (84.3%) and Fable 5 (83.4%)
SWE-bench Pro64.6%#6 on public leaderboard; K3 not yet listed
Agents' Last Exam (ALE)53.6+13.1 vs Fable 5 (adaptive); K3 not scored
AA Coding Agent Index80.0+2.8 vs Fable 5; uses fewer tokens at lower cost
BrowseComp (Ultra)92.2%Above K3's 91.2%
ExploitBench73.5%Cybersecurity; K3 not scored
OSWorld 2.062.6%Computer use; K3 not scored
HealthBench Professional60.5%Clinical reasoning; K3 not scored

Sol's Ultra mode is a genuine differentiator. At 91.9% on Terminal-Bench 2.1, it's the highest published score on that benchmark. The ALE score of 53.6 is 13.1 points above Fable 5. And Sol is notably token-efficient: Artificial Analysis measured ~15,000 output tokens per Intelligence Index task vs K3's much higher output volume, meaning Sol's higher per-token price is partially offset by using fewer tokens.

Radar: six shared benchmarks

Term-BenchDeepSWEFrontierSWEBrowseCompGDPval-AAGPQA Diam.

● Kimi K3 ● GPT-5.6 Sol

The radar uses six benchmarks where both models have scores. Normalized: Term-Bench, DeepSWE, FrontierSWE, BrowseComp, GPQA to 100; GDPval-AA to 2000. It is a visual aid, not a new benchmark.

  • Sol leads on Terminal-Bench 2.1, DeepSWE, GDPval-AA, and GPQA Diamond.
  • K3 leads on FrontierSWE and BrowseComp.
  • The radar shows how close these models are — the polygons nearly overlap on most axes.
  • Sol's Ultra mode (91.9% Terminal-Bench) and ALE score (53.6) are not shown here but represent additional capability headroom.

Pricing: K3 costs 40% less, but Sol is more token-efficient

Per 1M tokensKimi K3GPT-5.6 Sol
Fresh input / cache miss$3.00$5.00
Cached input / cache hit$0.30$0.50
Output$15.00$30.00
10M fresh input + 2M output$60.00$110.00
10M cached input + 2M output$33.00$65.00
Standard workload: 10M input + 2M output
K3 official pricing, OpenAI official pricing. Sol >272K input triggers 2× input / 1.5× output multipliers.
K3
$60
Sol
$110

K3 is cheaper on every token class. But the real cost story is more nuanced. Artificial Analysis found K3 used roughly 130 million output tokens across its Intelligence Index evaluation, compared with Sol's ~70 million — nearly double. K3's output tokens cost half as much, but it used almost twice as many. Total evaluation spending ended up in a similar range. Sol is also more token-efficient per-task: ~15,000 output tokens per Intelligence Index task vs higher counts for K3.

The practical takeaway: cost per accepted task, not cost per token. K3's always-on max reasoning can consume more output tokens than expected. Sol's adjustable reasoning lets you spend less on routine work. Measure both on your actual workload.

Ecosystem & deployment

K3: open weights coming, API-only today

K3's full weights are promised by July 27 under a Modified MIT license. At 2.8T parameters with 16 active experts, self-hosting requires at least 64 accelerators. Until then, K3 is API-only through Moonshot's platform and Kimi products (Kimi.com, Kimi Work, Kimi Code).

The open-weight promise enables fine-tuning, air-gapped deployment, and sovereignty that closed models can't match. But the infrastructure burden is real — most teams will use an inference provider even after the weights land.

Sol: closed, but deeply integrated

Sol is available through ChatGPT, Codex, and the OpenAI API. It's part of a three-tier family: Sol (flagship, $5/$30), Terra (balanced, $2.50/$15), and Luna (fastest, $1/$6). This tiering lets you route simple tasks to cheaper models and reserve Sol for hard problems.

Sol also runs on Cerebras at up to 750 tok/s for select customers — a latency play for interactive agents. The OpenAI ecosystem (Codex CLI, web search, file search, computer use) is more mature than Moonshot's tooling.

Which model should you use?

Choose Kimi K3 if…

  • You need the #1 frontend coding model — K3 leads the Frontend Code Arena.
  • Your workload is long-horizon, terminal-heavy, or browse-intensive — K3 wins FrontierSWE (+9.9), BrowseComp, and AutomationBench.
  • API cost is a first-order constraint — K3 is 40% cheaper at list price.
  • You need video input, air-gapped deployment, or future open-weight flexibility.
  • You want to run high-volume, measurable coding tasks with clear pass/fail criteria.

Choose GPT-5.6 Sol if…

  • You need maximum coding capability — Sol leads DeepSWE (+5.5), SWE-bench Pro (64.6%), and Terminal-Bench 2.1.
  • You want Ultra mode for the hardest problems — 91.9% Terminal-Bench 2.1 with parallel sub-agents.
  • You value adjustable reasoning — dial effort from none through max per task.
  • You rely on OpenAI's ecosystem: Codex, web search, file search, computer use, Cerebras speed.
  • You want the GPT-5.6 family's tiered pricing — route simple tasks to Terra or Luna.
Recommended routing: use K3 as the high-volume default for frontend generation, terminal agents, long-horizon coding, and web research — tasks where its benchmark wins and 40% price advantage translate to real savings. Route difficult repository engineering, controlled reasoning, and high-stakes production work to Sol. For the hardest problems, Sol Ultra is in a class of its own. The strongest architecture uses both: K3 for volume and open-weight flexibility, Sol for ceiling and ecosystem depth.

Compare both models on real code

Run the same repository prompt, tool schema, and acceptance tests rather than trusting a leaderboard headline.

Try both in CodingFleet →

Final verdict

GPT-5.6 Sol is the stronger coding model. It leads on 6 of 9 shared benchmarks, posts the highest Terminal-Bench 2.1 score in Ultra mode (91.9%), and brings a mature ecosystem with adjustable reasoning, tiered pricing across the GPT-5.6 family, and deep Codex integration. For difficult repository engineering and controlled reasoning, it's the safer choice.

Kimi K3 is the more disruptive product. It wins on FrontierSWE (+9.9), BrowseComp, AutomationBench, and AA-Briefcase. It's #1 on the Frontend Code Arena. It costs 40% less at list price. And it will soon be available as open weights — a strategic advantage no closed model can match. For frontend coding, long-horizon agents, and cost-sensitive pipelines, it's extraordinarily compelling.

The honest assessment: Sol is the better model for hard engineering. K3 is the better deal for high-volume coding work. The gap between them is narrow enough — and the strengths are different enough — that the smartest architecture routes by task, not by model. Test both on your actual workload.

Sources and methodology