Archive · 2026-10

October 2026,
frozen.

These figures were written once, at month close on 1 Nov 2026, and cannot be edited. Cite them freely: this page will read the same in a year.

What this is: a public monthly record of what AI model prices actually did, which providers moved them, and which model is still the cheapest one good enough for a given job.

510

Models tracked

Active entries in the live catalog.

80

Providers tracked

Distinct real providers with at least one live price. Aggregator listings are excluded.

19,117

Price moves in October 2026

8062 up · 11055 down. 232 new listings are counted separately.

Number of the month

+24650%

Qwen3.8 27B, same weights

$0.0040 at PrimeIntellect against $0.990 at Cerebras, per MTok in.

Direction

19117 recorded price moves in October 2026: 11055 cuts and 8062 rises, so the market moved down on balance. 58% of all moves were cuts.

Four terms this page uses

MTok
One million tokens. Every price on this page is US dollars per million tokens, so models with different token accounting stay comparable.
Price move
One observed change to a live listed price, recorded in the append-only ledger with its direction. A model appearing for the first time is a new listing, never a move. Full definition.
Provider spread
The gap between the cheapest and dearest host serving identical weights. It is a routing choice, not a quality difference.
Equivalence band
The score range around your current model, widened by the benchmark's own measurement margin, that a candidate must sit inside before a switch is offered.

Price moves

What changed in October 2026

Recorded from the append-only price ledger: every observed move, per model and per host, on both the input and output side.

19117price movesOctober 2026
  • Decreases11055
  • Increases8062
  • New listings232
  • Total moves counts increases plus decreases only. New listings are shown here for context and are never folded into that total.
19,117

Total moves

Increases plus decreases. New listings are not folded in.

8,062

Increases

11,055

Decreases

34

New models

First seen in October 2026.

Top 5 increases

  • deepseek/deepseek-v4-flash-0731

    OpenInference

    +9386.8%

    blended

    $0.0041 → $0.011 in (+168.6%) · $0.0095 → $1.28 out (+13334.1%)

  • moonshotai/kimi-k3

    OpenRouter

    +8340.0%

    blended

    $0.030 → $0.192 in (+540.0%) · $0.150 → $15.00 out (+9900.0%)

  • ~deepseek/deepseek-v4-flash-latest

    OpenRouter

    +8205.4%

    blended

    $0.0049 → $0.0046 in (-5.7%) · $0.011 → $1.28 out (+11990.3%)

  • ~deepseek/deepseek-v4-flash-latest

    OpenRouter

    +7925.6%

    blended

    $0.0064 → $0.0061 in (-4.7%) · $0.0096 → $1.28 out (+13198.7%)

  • ~deepseek/deepseek-v4-flash-latest

    OpenRouter

    +6392.2%

    blended

    $0.0051 → $0.0048 in (-6.3%) · $0.015 → $1.28 out (+8625.3%)

Top 5 decreases

  • moonshotai/kimi-k3

    OpenRouter

    -99.0%

    blended

    $0.239 → $0.0024 in (-99.0%) · $10.80 → $0.108 out (-99.0%)

  • ~moonshotai/kimi-latest

    OpenRouter

    -98.9%

    blended

    $0.162 → $0.0024 in (-98.5%) · $9.75 → $0.108 out (-98.9%)

  • ~deepseek/deepseek-v4-flash-latest

    OpenRouter

    -98.9%

    blended

    $0.0046 → $0.0037 in (-20.0%) · $1.28 → $0.011 out (-99.2%)

  • ~deepseek/deepseek-v4-flash-latest

    OpenRouter

    -98.8%

    blended

    $0.0057 → $0.0046 in (-20.0%) · $1.28 → $0.011 out (-99.2%)

  • ~deepseek/deepseek-v4-flash-latest

    OpenRouter

    -98.8%

    blended

    $0.0061 → $0.0049 in (-20.0%) · $1.28 → $0.011 out (-99.2%)

Who reprices most

Ranked over everything we hold. The leaderboard updates as more price moves are recorded.

  • 1OpenRouter9044moves · 90 models
  • 2Relace2078moves · 11 models
  • 3Morph1596moves · 7 models
  • 4OpenInference1515moves · 6 models
  • 5Wafer1290moves · 11 models
  • 6Baidu715moves · 11 models
  • 7InferenceNet480moves · 9 models
  • 8Inceptron458moves · 6 models

Market structure

The same weights cost wildly different money

Identical model, different real provider. Aggregator listings are excluded, so every gap below is a genuine provider-to-provider spread you could act on today.

143

Models on 2+ providers

1

Median providers per model

37

Most providers on one model

257
164%
79
2–320%
42
4–911%
22
10+6%

Providers per model, across 400 models with at least one real (non-aggregator) endpoint. Most weights are single-sourced; a small tail is served everywhere, and that tail is where the spread lives.

  • Qwen3.8 27B

    qwen/qwen3.8-27b

    $0.0040$0.990

    PrimeIntellect → Cerebras · 22 providers · per MTok in

    +24650%

    spread

  • GLM 5.2

    z-ai/glm-5.2

    $0.0060$1.40

    Wafer → GMICloud · 25 providers · per MTok in

    +23233%

    spread

  • DeepSeek V4.1 Flash

    deepseek/deepseek-v4.1-flash

    $0.011$0.300

    Relace → Google · 32 providers · per MTok in

    +2579%

    spread

  • DeepSeek V4.1 Flash 1017

    deepseek/deepseek-v4.1-flash-1017

    $0.012$0.300

    OpenInference → Parasail · 35 providers · per MTok in

    +2442%

    spread

  • GLM 5.3

    z-ai/glm-5.3

    $0.070$1.68

    Wafer → Amazon Bedrock · 37 providers · per MTok in

    +2303%

    spread

  • Kimi K3

    moonshotai/kimi-k3

    $0.149$3.45

    Sail Research → Alibaba · 21 providers · per MTok in

    +2208%

    spread

  • DeepSeek V4 Pro 0813

    deepseek/deepseek-v4-pro-0813

    $0.103$1.65

    Relace → Venice · 23 providers · per MTok in

    +1494%

    spread

  • DeepSeek V4 Flash 0423

    deepseek/deepseek-v4-flash

    $0.038$0.600

    Relace → Wafer · 17 providers · per MTok in

    +1483%

    spread

Quality per dollar

The cheapest model that still clears the band

For each task class we take the leading published score, subtract that evaluation's measured margin, and pick the cheapest model still above the line. Only suites with a real measured margin appear, because a benchmark without one cannot back a claim.

  • gpqa

    GLM 5.3 Flash

    z-ai/glm-5.3-flash

    Scores 91.20 on aa:gpqa; bar 91.10 (leader 96.10 − margin ±5.00). 31 models clear it. Cheapest listing at Relace.

    cheapest qualifying 91.20bar 91.10leader 96.10band width ±5.00 (measured margin)
    $0.015

    per MTok in

  • hle

    Claude Opus 5.5

    anthropic/claude-opus-5.5

    Scores 61.40 on aa:hle; bar 59.70 (leader 61.40 − margin ±1.70). 1 model clear it. Cheapest listing at Amazon Bedrock.

    cheapest qualifying 61.40bar 59.70leader 61.40band width ±1.70 (measured margin)
    $4.00

    per MTok in

  • lcr

    DeepSeek V4.1 Flash

    deepseek/deepseek-v4.1-flash

    Scores 84.00 on aa:lcr; bar 83.67 (leader 88.67 − margin ±5.00). 10 models clear it. Cheapest listing at Relace.

    cheapest qualifying 84.00bar 83.67leader 88.67band width ±5.00 (measured margin)
    $0.011

    per MTok in

  • scicode

    Claude Opus 5.5

    anthropic/claude-opus-5.5

    Scores 66.90 on aa:scicode; bar 61.90 (leader 66.90 − margin ±5.00). 2 models clear it. Cheapest listing at Amazon Bedrock.

    cheapest qualifying 66.90bar 61.90leader 66.90band width ±5.00 (measured margin)
    $4.00

    per MTok in

  • tau_banking

    Qwen3.8 27B

    qwen/qwen3.8-27b

    Scores 48.04 on aa:tau_banking; bar 45.72 (leader 50.72 − margin ±5.00). 9 models clear it. Cheapest listing at PrimeIntellect.

    cheapest qualifying 48.04bar 45.72leader 50.72band width ±5.00 (measured margin)
    $0.0040

    per MTok in

  • terminalbench_v2_1

    Gemini 3.8 Flash

    google/gemini-3.8-flash

    Scores 87.64 on aa:terminalbench_v2_1; bar 86.39 (leader 91.39 − margin ±5.00). 8 models clear it. Cheapest listing at Google.

    cheapest qualifying 87.64bar 86.39leader 91.39band width ±5.00 (measured margin)
    $0.375

    per MTok in

Benchmark saturation

An evaluation stops being usable when the spread between models collapses into the measurement margin. We require the observed spread to exceed twice the margin; a ratio at or below 1.0 means the instrument can no longer tell models apart.

  • tau_banking · aa:tau_banking

    spread 49.69 vs margin ±8.42 across 88 scored models, discriminating.

    2.95×vs 1.0× floor
  • terminalbench_v2_1 · aa:terminalbench_v2_1

    spread 88.02 vs margin ±10.32 across 89 scored models, discriminating.

    4.26×vs 1.0× floor
  • lcr · aa:lcr

    spread 86.67 vs margin ±9.39 across 161 scored models, discriminating.

    4.61×vs 1.0× floor
  • scicode · aa:scicode

    spread 49.50 vs margin ±5.33 across 175 scored models, discriminating.

    4.64×vs 1.0× floor
  • gpqa · aa:gpqa

    spread 61.20 vs margin ±5.83 across 155 scored models, discriminating.

    5.25×vs 1.0× floor
  • hle · aa:hle

    spread 59.60 vs margin ±1.70 across 175 scored models, discriminating.

    17.57×vs 1.0× floor

Newsletter

These numbers, weekly.

One email a week: the price moves we recorded, which switches cleared quality, and what changed for the cheapest model good enough for a given job. No pitch, unsubscribe in one click.

Method

How a switch gets measured

  • Price sync at freeze

    The prices on this page are the catalog as it stood when this month was frozen. The engine itself keeps re-syncing from public provider feeds; the live page reflects that, this archive deliberately does not.

  • Independent benchmark scores

    Quality comes from published third-party evaluations, per task class. We do not run our own private eval and we are never paid for placement.

  • The equivalence band

    A candidate model only qualifies when its score sits inside the band around your current model for the task class in question. Cheaper-but-worse never clears.

  • Measurement margin

    Every score carries its own uncertainty. We compute the real margin and require the gap to survive it before a switch is offered.

  • Latency ceilings

    Median latency is part of the decision, not an afterthought. Set a ceiling and candidates that breach it are dropped before cost is even compared.

  • Refusals with reasons

    When nothing clears, you get a stated reason, not a weaker suggestion. A quiet downgrade would cost you more than the saving is worth.

Cite and reuse

Take the numbers with you

Every figure here is free to quote, repost or chart, with attribution to CostMyAI. We would rather be the cited source than the hidden one.

Citation

CostMyAI Intelligence, October 2026. Retrieved 1 Nov 2026. https://www.costmyai.com/intelligence/2026-10

Permanent link

https://www.costmyai.com/intelligence/2026-10?ref=share&card=cite

Download the table

Same figures, same identifiers as the anchors on this page, so a chart you build from the file matches the page it came from.

Every figure card on this page carries its own image and a ready post with the source line already in it.

Embed

Put the live market on your own page

A self-contained frame that rotates the three sharpest numbers of the month: how many prices moved, the steepest rise and the steepest cut. It refreshes itself, needs no script on your site, and always credits CostMyAI.

<iframe src="https://costmyai.com/embed/intelligence-widget"
  title="AI price market, via CostMyAI"
  width="100%" height="200" loading="lazy"
  style="border:0;max-width:520px"
  referrerpolicy="strict-origin-when-cross-origin"></iframe>
  • RotationMonth-over-month move count, biggest rise, biggest cut.
  • FreshnessServer-cached and refreshed every five minutes.
  • SafetyIsolated iframe, no script in your page, nothing configurable.
  • AttributionThe “via CostMyAI” link is part of the widget on every plan.

Archive

Every closed month, frozen and permanently linkable

At 00:00 UTC on the first of each month we write that month's final figures once and never touch them again. A correction is filed as a new restatement row that points back at the original, so the number you cited stays exactly as you cited it.