Skip to content

How TreadLightlyAI compares

Side-by-side with the major chatbots on privacy, energy disclosure, and features.

Last updated 2026-10-29

Models in TreadLightlyAI

Five chat models, four engines, three providers, with blended energy rates per token. See our energy methodology for how every rate below was derived.

ModelCreatorUnderlying modelParametersEnergy rateData centerElectricityLast checked
Green TinyGoogle DeepMindGemma 4 31B Instruct (capped output)31B~108 µWh/token(estimated)Infomaniak (Geneva, CH)Certified renewable2026-08-03
Green SmallGoogle DeepMindGemma 4 31B Instruct31B~108 µWh/token(estimated)Infomaniak (Geneva, CH)Certified renewable2026-08-03
Green LargeOpenAIGPT-OSS 120B120B115 µWh/token(measured)GreenPT (data center not named)Certified renewable, per GreenPT2026-04-20
HaikuAnthropicClaude Haiku 5.5Not disclosed~78 µWh/token(estimated)Anthropic (not disclosed)Not disclosed2026-10-11
SonnetAnthropicClaude Sonnet 5.5Not disclosed~240 µWh/token(estimated)Anthropic (not disclosed)Not disclosed2026-10-11

Energy rates are blended: one rate covers tokens you send and tokens the model writes. Green Large's rate is measured from GreenPT per-call data; every other rate, including Green Tiny and Green Small, is estimated (methodology version ). Estimated rates include 20% for cooling, networking, and storage. Green Tiny and Green Small share one engine (Gemma 4 31B at Infomaniak) and carry the same rate. Our estimate for Claude Haiku is lower than our estimate for Green Tiny and Green Small. The green models are chosen for where their electricity comes from and who trained the model, not for using less of it than everything else on offer. Full methodology

Transparency commitments

What each service does and discloses about energy, data, and safety.

DimensionTreadLightlyAIChatGPT(opens in new tab)Gemini(opens in new tab)Copilot(opens in new tab)Claude(opens in new tab)
Shows energy per conversationYesNoNoNoNo
Anti-sycophancy by designYesPartialNoNoPartial
Funds AI safety and carbon removal10% of all revenue to FLI; 1.5% to carbon removalNoNoNoPartial
Published AI limitationsAI Transparency CardPartialPartialPartialYes
No user data trainingYesNoNoPartialPartial
Financial transparencyPartialNoPartialPartialNo
AI alternatives guidanceYesNoNoNoNo

Feature Comparison

What each service offers, as of 2026-10-29.

DimensionTreadLightlyAIChatGPT(opens in new tab)Gemini(opens in new tab)Copilot(opens in new tab)Claude(opens in new tab)
Large frontier modelsNo (intentional)Yes (flagged)GPT-5.4Yes (flagged)Gemini 2.5 ProYes (flagged)GPT-5.4Yes (flagged)Opus
Medium modelsGreen Large; Haiku and Sonnet on AdvancedGPT-4o-miniGemini 2.0 FlashGPT-4o-miniSonnet, Haiku
Small modelsGreen Small, Green TinyNoFlash LiteNoNo
Code assistanceBasicYesYesYesYes
Thinking modesYesPartialPartialPartialPartial
Prompt improvementRefine and rewrite optionsNoNoNoNo
Project foldersYesYesGemsPartialYes
Conversation history & exportYesYesYesPartialPartial
Image & document analysisAll tiers, green models includedYesYesYesYes
Image generationNo (intentional)Yes (flagged)Yes (flagged)Yes (flagged)No (intentional)
Web searchYesYesYesYesYes
Enterprise featuresNoYesYesYesYes
Free tierYesYesYesYesYes
Pricing (paid tiers)$3, 8, 20 /mo$20, 200 /mo$20/mo$20/mo$20, 100, 200 /mo

Think first. Prompt second.

Get started free. No credit card required.

Get started

If TreadLightlyAI is not a fit, you have other good choices.

We maintain a comparison of responsible AI chatbots at Dangerous Robot: Responsible AI(opens in new tab)