Skip to content
Signals
deepseek-ai/deepseek-harness adds ~58,590 stars this week — breakout in TypeScriptTrendNexus signalsstorytold/pdfcraft adds ~27,969 stars this week — emerging in RustTrendNexus signalsEbony-Vinyl/dsh-our-free-model adds ~20,064 stars this week — emerging in JavaScriptTrendNexus signalseternity4719/HowToLiveBetter adds ~19,959 stars this week — breakout in HTMLTrendNexus signalsgarrytan/gstack adds ~15,268 stars this week — breakout in TypeScriptTrendNexus signalsGraphify-Labs/graphify adds ~15,150 stars this week — breakout in PythonTrendNexus signalsodysseus-dev/odysseus adds ~12,447 stars this week — breakout in PythonTrendNexus signalsPanniantong/Agent-Reach adds ~12,224 stars this week — breakout in PythonTrendNexus signalsRisk watch: bagisto/bagisto trends cooling with security score 92TrendNexus signalsRisk watch: thedaviddias/Front-End-Checklist trends cooling with security score 82TrendNexus signalsHealthiest in cohort: ultraworkers/claw-code holds PVI 84 — sustainability leaderTrendNexus signals

Kimi K3

Moonshot · Open agentic muscle.

Kimi's K-line is the open ecosystem's agentic specialist — trained hard on tool use and long-horizon tasks, at value pricing.

Live OpenRouter pricing · capability ratings & benchmarks are indicative

Context window

1.0M

tokens

Input · $/1M

$0.73

per million tokens

Output · $/1M

$10.00

per million tokens

Vs. cost floor

16×

output cost vs DeepSeek V4.1 Flash

Capability profile

Editorial ratings across six axes (0–100), compiled from publisher reports and public evals — indicative, not measured by TrendNexus.

Benchmark highs

SWE-bench Verified≈68%

strong open-agentic result

Tau-bench (tools)≈75%

tool-use reliability

AIME-class math≈88%

solid math

MMLU-Pro≈81%

broad knowledge

Where it sits in the field

Output cost per million tokens across all 15 tracked flagships — Kimi K3 highlighted.

Claude Fable 5.1$50.00GPT-5.5$30.00Claude Opus 5.5$20.00GPT-5.6 Terra Pro$12.00Gemini 3.1 Pro Preview$12.00GPT-6.1 Sol Pro$10.00Claude Sonnet 5.5$10.00Kimi K3$10.00GLM 5.3 Prime$8.80Grok 4.7$6.00Qwen3 Max$6.00Gemini 3.8 Flash$3.75DeepSeek V4 Pro 0423$0.872Llama 4 Maverick$0.652DeepSeek V4.1 Flash$0.621

What it's suited to build

Product classes where this model's profile wins — mockups are illustrative, not model output.

Agents

Open-stack agent systems

Agentic backbones without closed-model dependency.

Coding

Automation engineering

Scripted operations and integration glue at low cost.

Self-hosted / edge

Hybrid deployments

API or self-hosted as economics dictate.

Benchmark figures and capability ratings are indicative — compiled from publisher reports and public leaderboards, refreshed editorially, and not independently measured by TrendNexus. Verify against your own evals before procurement decisions. Pricing updates with the daily build.