---
title: "Context cliffs — every tier step function — Undominated"
canonical: https://undominated.ai/cliffs/
description: "71 models change rate past a prompt threshold; the steepest published ladder reaches 6.67× on input (as of 2026-10-06)."
---

# Context cliffs — every tier step function — Undominated

> 71 models change rate past a prompt threshold; the steepest published ladder reaches 6.67× on input (as of 2026-10-06).

# Context cliffs

Every published context-tier step function in the catalogue. Ladders keep every rung — a flattened ladder once mispriced five models by up to 2×.

As of 2026-10-06, **71** publishable models raise the input rate past a prompt-length threshold. The steepest published ladder reaches **6.67×** on input (Qwen3.7 Flash) , conditional on the prompt crossing that model’s highest published rung — not on the headline rate.

5 of 71 carry more than one rung. Multipliers are input $/M at the rung ÷ headline input $/M. Condition: OpenRouter-published tiers in the catalogue payload.

| Model | Headline in | Rungs | Max × |
| --- | --- | --- | --- |
| [Qwen3.7 Flash](/models/qwen__qwen3.7-flash/) [Check](/check/qwen__qwen3.7-flash/) > 32K → $0.100/M in (3.33×) first rung, not the top > 256K → $0.200/M in (6.67×) | $0.030 | 2 | 6.67× |
| [Qwen3.6 Flash](/models/qwen__qwen3.6-flash/) [Check](/check/qwen__qwen3.6-flash/) > 256K → $0.750/M in (4×) | $0.188 | 1 | 4× |
| [Qwen3.6 Plus](/models/qwen__qwen3.6-plus/) [Check](/check/qwen__qwen3.6-plus/) > 256K → $1.30/M in (4×) | $0.325 | 1 | 4× |
| [Qwen Plus 0728](/models/qwen__qwen-plus-2025-07-28/) [Check](/check/qwen__qwen-plus-2025-07-28/) > 256K → $0.780/M in (3×) | $0.260 | 1 | 3× |
| [Qwen-Plus](/models/qwen__qwen-plus/) [Check](/check/qwen__qwen-plus/) > 256K → $0.780/M in (3×) | $0.260 | 1 | 3× |
| [Qwen3 Coder Plus](/models/qwen__qwen3-coder-plus/) [Check](/check/qwen__qwen3-coder-plus/) > 32K → $1.17/M in (1.8×) first rung, not the top > 128K → $1.95/M in (3×) | $0.650 | 2 | 3× |
| [Qwen3.7 Plus](/models/qwen__qwen3.7-plus/) [Check](/check/qwen__qwen3.7-plus/) > 256K → $0.960/M in (3×) | $0.320 | 1 | 3× |
| [Qwen3 Coder Flash](/models/qwen__qwen3-coder-flash/) [Check](/check/qwen__qwen3-coder-flash/) > 32K → $0.325/M in (1.67×) first rung, not the top > 128K → $0.520/M in (2.67×) | $0.195 | 2 | 2.67× |
| [Qwen3 Max](/models/qwen__qwen3-max/) [Check](/check/qwen__qwen3-max/) > 32K → $1.56/M in (2×) first rung, not the top > 128K → $1.95/M in (2.5×) | $0.780 | 2 | 2.5× |
| [Qwen3 Max Thinking](/models/qwen__qwen3-max-thinking/) [Check](/check/qwen__qwen3-max-thinking/) > 32K → $1.56/M in (2×) first rung, not the top > 128K → $1.95/M in (2.5×) | $0.780 | 2 | 2.5× |
| [Claude Sonnet 4](/models/anthropic__claude-sonnet-4/) [Check](/check/anthropic__claude-sonnet-4/) > 200K → $6.00/M in (2×) | $3.00 | 1 | 2× |
| [Claude Sonnet 4.5](/models/anthropic__claude-sonnet-4.5/) [Check](/check/anthropic__claude-sonnet-4.5/) > 200K → $6.00/M in (2×) | $3.00 | 1 | 2× |
| [Claude Sonnet 4.5 (batch)](/models/anthropic__claude-sonnet-4.5__batch/) > 200K → $3.00/M in (2×) | $1.50 | 1 | 2× |
| [Fugu Ultra](/models/sakana__fugu-ultra/) [Check](/check/sakana__fugu-ultra/) > 272K → $10.00/M in (2×) | $5.00 | 1 | 2× |
| [Fugu Ultra v2](/models/sakana__fugu-ultra-v2/) [Check](/check/sakana__fugu-ultra-v2/) > 272K → $10.00/M in (2×) | $5.00 | 1 | 2× |
| [Gemini 2.5 Pro](/models/google__gemini-2.5-pro/) [Check](/check/google__gemini-2.5-pro/) > 200K → $2.50/M in (2×) | $1.25 | 1 | 2× |
| [Gemini 2.5 Pro (batch)](/models/google__gemini-2.5-pro__batch/) > 200K → $1.25/M in (2×) | $0.625 | 1 | 2× |
| [Gemini 2.5 Pro Preview 06-05](/models/google__gemini-2.5-pro-preview/) [Check](/check/google__gemini-2.5-pro-preview/) > 200K → $2.50/M in (2×) | $1.25 | 1 | 2× |
| [Gemini 3.1 Pro Preview](/models/google__gemini-3.1-pro-preview/) [Check](/check/google__gemini-3.1-pro-preview/) > 200K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [Gemini 3.1 Pro Preview (batch)](/models/google__gemini-3.1-pro-preview__batch/) > 200K → $2.00/M in (2×) | $1.00 | 1 | 2× |
| [Gemini 3.1 Pro Preview Custom Tools](/models/google__gemini-3.1-pro-preview-customtools/) [Check](/check/google__gemini-3.1-pro-preview-customtools/) > 200K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [GPT-5.4](/models/openai__gpt-5.4/) [Check](/check/openai__gpt-5.4/) > 272K → $5.00/M in (2×) | $2.50 | 1 | 2× |
| [GPT-5.4 (batch)](/models/openai__gpt-5.4__batch/) > 272K → $2.50/M in (2×) | $1.25 | 1 | 2× |
| [GPT-5.4 Pro](/models/openai__gpt-5.4-pro/) [Check](/check/openai__gpt-5.4-pro/) > 272K → $60.00/M in (2×) | $30.00 | 1 | 2× |
| [GPT-5.4 Pro (batch)](/models/openai__gpt-5.4-pro__batch/) > 272K → $30.00/M in (2×) | $15.00 | 1 | 2× |
| [GPT-5.5](/models/openai__gpt-5.5/) [Check](/check/openai__gpt-5.5/) > 272K → $10.00/M in (2×) | $5.00 | 1 | 2× |
| [GPT-5.5 (batch)](/models/openai__gpt-5.5__batch/) > 272K → $5.00/M in (2×) | $2.50 | 1 | 2× |
| [GPT-5.5 Pro](/models/openai__gpt-5.5-pro/) [Check](/check/openai__gpt-5.5-pro/) > 272K → $60.00/M in (2×) | $30.00 | 1 | 2× |
| [GPT-5.5 Pro (batch)](/models/openai__gpt-5.5-pro__batch/) > 272K → $30.00/M in (2×) | $15.00 | 1 | 2× |
| [GPT-5.6 Luna](/models/openai__gpt-5.6-luna/) [Check](/check/openai__gpt-5.6-luna/) > 272K → $0.400/M in (2×) | $0.200 | 1 | 2× |
| [GPT-5.6 Luna (batch)](/models/openai__gpt-5.6-luna__batch/) > 272K → $0.200/M in (2×) | $0.100 | 1 | 2× |
| [GPT-5.6 Luna Pro](/models/openai__gpt-5.6-luna-pro/) [Check](/check/openai__gpt-5.6-luna-pro/) > 272K → $0.400/M in (2×) | $0.200 | 1 | 2× |
| [GPT-5.6 Luna Pro (batch)](/models/openai__gpt-5.6-luna-pro__batch/) > 272K → $0.200/M in (2×) | $0.100 | 1 | 2× |
| [GPT-5.6 Sol](/models/openai__gpt-5.6-sol/) [Check](/check/openai__gpt-5.6-sol/) > 272K → $8.00/M in (2×) | $4.00 | 1 | 2× |
| [GPT-5.6 Sol (batch)](/models/openai__gpt-5.6-sol__batch/) > 272K → $2.00/M in (2×) | $1.00 | 1 | 2× |
| [GPT-5.6 Sol Pro](/models/openai__gpt-5.6-sol-pro/) [Check](/check/openai__gpt-5.6-sol-pro/) > 272K → $8.00/M in (2×) | $4.00 | 1 | 2× |
| [GPT-5.6 Sol Pro (batch)](/models/openai__gpt-5.6-sol-pro__batch/) > 272K → $2.00/M in (2×) | $1.00 | 1 | 2× |
| [GPT-5.6 Terra](/models/openai__gpt-5.6-terra/) [Check](/check/openai__gpt-5.6-terra/) > 272K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [GPT-5.6 Terra (batch)](/models/openai__gpt-5.6-terra__batch/) > 272K → $2.00/M in (2×) | $1.00 | 1 | 2× |
| [GPT-5.6 Terra Pro](/models/openai__gpt-5.6-terra-pro/) [Check](/check/openai__gpt-5.6-terra-pro/) > 272K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [GPT-5.6 Terra Pro (batch)](/models/openai__gpt-5.6-terra-pro__batch/) > 272K → $2.00/M in (2×) | $1.00 | 1 | 2× |
| [GPT-6 Astra](/models/openai__gpt-6-astra/) [Check](/check/openai__gpt-6-astra/) > 272K → $20.00/M in (2×) | $10.00 | 1 | 2× |
| [GPT-6 Astra (batch)](/models/openai__gpt-6-astra__batch/) > 272K → $10.00/M in (2×) | $5.00 | 1 | 2× |
| [GPT-6 Astra Pro](/models/openai__gpt-6-astra-pro/) [Check](/check/openai__gpt-6-astra-pro/) > 272K → $20.00/M in (2×) | $10.00 | 1 | 2× |
| [GPT-6 Astra Pro (batch)](/models/openai__gpt-6-astra-pro__batch/) > 272K → $10.00/M in (2×) | $5.00 | 1 | 2× |
| [GPT-6 Luna](/models/openai__gpt-6-luna/) [Check](/check/openai__gpt-6-luna/) > 272K → $0.200/M in (2×) | $0.100 | 1 | 2× |
| [GPT-6 Luna (batch)](/models/openai__gpt-6-luna__batch/) > 272K → $0.100/M in (2×) | $0.050 | 1 | 2× |
| [GPT-6 Luna Pro](/models/openai__gpt-6-luna-pro/) [Check](/check/openai__gpt-6-luna-pro/) > 272K → $0.200/M in (2×) | $0.100 | 1 | 2× |
| [GPT-6 Luna Pro (batch)](/models/openai__gpt-6-luna-pro__batch/) > 272K → $0.100/M in (2×) | $0.050 | 1 | 2× |
| [GPT-6 Sol](/models/openai__gpt-6-sol/) [Check](/check/openai__gpt-6-sol/) > 272K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [GPT-6 Sol (batch)](/models/openai__gpt-6-sol__batch/) > 272K → $2.00/M in (2×) | $1.00 | 1 | 2× |
| [GPT-6 Sol Pro](/models/openai__gpt-6-sol-pro/) [Check](/check/openai__gpt-6-sol-pro/) > 272K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [GPT-6 Sol Pro (batch)](/models/openai__gpt-6-sol-pro__batch/) > 272K → $2.00/M in (2×) | $1.00 | 1 | 2× |
| [GPT-6.1 Sol](/models/openai__gpt-6.1-sol/) [Check](/check/openai__gpt-6.1-sol/) > 272K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [GPT-6.1 Sol Pro](/models/openai__gpt-6.1-sol-pro/) [Check](/check/openai__gpt-6.1-sol-pro/) > 272K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [Grok 4.20](/models/x-ai__grok-4.20/) [Check](/check/x-ai__grok-4.20/) > 200K → $2.50/M in (2×) | $1.25 | 1 | 2× |
| [Grok 4.20 Multi-Agent](/models/x-ai__grok-4.20-multi-agent/) [Check](/check/x-ai__grok-4.20-multi-agent/) > 200K → $2.50/M in (2×) | $1.25 | 1 | 2× |
| [Grok 4.3](/models/x-ai__grok-4.3/) [Check](/check/x-ai__grok-4.3/) > 200K → $2.50/M in (2×) | $1.25 | 1 | 2× |
| [Grok 4.3 (batch)](/models/x-ai__grok-4.3__batch/) > 200K → $2.00/M in (2×) | $1.00 | 1 | 2× |
| [Grok 4.5](/models/x-ai__grok-4.5/) [Check](/check/x-ai__grok-4.5/) > 200K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [Grok 4.6](/models/x-ai__grok-4.6/) [Check](/check/x-ai__grok-4.6/) > 200K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [Grok 4.7](/models/x-ai__grok-4.7/) [Check](/check/x-ai__grok-4.7/) > 200K → $4.00/M in (2×) | $2.00 | 1 | 2× |
| [Grok Build 0.1](/models/x-ai__grok-build-0.1/) [Check](/check/x-ai__grok-build-0.1/) > 200K → $2.00/M in (2×) | $1.00 | 1 | 2× |
| [Seed 1.6](/models/bytedance-seed__seed-1.6/) [Check](/check/bytedance-seed__seed-1.6/) > 128K → $0.500/M in (2×) | $0.250 | 1 | 2× |
| [Seed-2.0-Code](/models/bytedance-seed__seed-2.0-code/) [Check](/check/bytedance-seed__seed-2.0-code/) > 128K → $1.00/M in (2×) | $0.500 | 1 | 2× |
| [Seed-2.0-Lite](/models/bytedance-seed__seed-2.0-lite/) [Check](/check/bytedance-seed__seed-2.0-lite/) > 128K → $0.500/M in (2×) | $0.250 | 1 | 2× |
| [Seed-2.0-Mini](/models/bytedance-seed__seed-2.0-mini/) [Check](/check/bytedance-seed__seed-2.0-mini/) > 128K → $0.200/M in (2×) | $0.100 | 1 | 2× |
| [Qwen3.6 Max Preview](/models/qwen__qwen3.6-max-preview/) [Check](/check/qwen__qwen3.6-max-preview/) > 128K → $1.58/M in (1.54×) | $1.03 | 1 | 1.54× |
| [Seed 1.6 Flash](/models/bytedance-seed__seed-1.6-flash/) [Check](/check/bytedance-seed__seed-1.6-flash/) > 128K → $0.100/M in (1.33×) | $0.075 | 1 | 1.33× |
| [Qwen3.5 Plus 2026-02-15](/models/qwen__qwen3.5-plus-02-15/) [Check](/check/qwen__qwen3.5-plus-02-15/) > 256K → $0.325/M in (1.25×) | $0.260 | 1 | 1.25× |
| [Qwen3.5 Plus 2026-04-20](/models/qwen__qwen3.5-plus-20260420/) [Check](/check/qwen__qwen3.5-plus-20260420/) > 256K → $0.375/M in (1.25×) | $0.300 | 1 | 1.25× |

## Questions this page answers

 What is a context cliff?

A published step where input $/M rises after a prompt-length threshold. The headline rate applies through the first threshold.

 Why does every rung stay visible?

Flattening a tier ladder to one multiplier once mispriced five models by up to 2×. Each rung is kept so long prompts are priced correctly.

 Are output tiers included?

This table lists input-tier ladders from the OpenRouter-published payload. Output-tier cliffs appear when vendors publish them.

 Copy evidence Download CSV · 76 rows [Input hashes & method version](/data/publications/manifests/ecef83f767cd4d2c21766f6a5b4dffa36c06351f23657542d473bc1a030c0932.json)

 Read citation text and BibTeX 71 model entries have published context tiers; 5 preserve multiple rungs. Tier rates apply strictly above each threshold; missing cache prices remain unknown. As of 2026-10-06. Lens: Published provider prices in the accepted OpenRouter-derived catalogue. https://undominated.ai/cliffs/
Method: https://undominated.ai/methodology/
Publication: sha256:ecef83f767cd4d2c21766f6a5b4dffa36c06351f23657542d473bc1a030c0932; calculation SHA-256: 5ae4d0bee9b70298dd437b05df87c32acb407040b50bf88240677f295f25955d; permitted inputs: https://undominated.ai/data/publications/manifests/ecef83f767cd4d2c21766f6a5b4dffa36c06351f23657542d473bc1a030c0932.json

Copy BibTeX Download citation

 @misc{undominated-cliffs-2026-10-06,
 author = {{Undominated.ai}},
 title = {cliffs evidence — Undominated.ai},
 year = {2026},
 url = {https://undominated.ai/cliffs/},
 note = {as of 2026-10-06; Published provider prices in the accepted OpenRouter-derived catalogue},
 annote = {Publication: sha256:ecef83f767cd4d2c21766f6a5b4dffa36c06351f23657542d473bc1a030c0932; calculation SHA-256: 5ae4d0bee9b70298dd437b05df87c32acb407040b50bf88240677f295f25955d; permitted inputs: https://undominated.ai/data/publications/manifests/ecef83f767cd4d2c21766f6a5b4dffa36c06351f23657542d473bc1a030c0932.json}
}

## Continue your investigation

 - [Traps](/traps/)
- [Spreads](/spreads/)
- [Cache](/cache/)
- [Methodology](/methodology/)