DEV Community: 4663437Mehdi

The Token Ledger – 2026-08-06

4663437Mehdi — Thu, 06 Aug 2026 09:50:33 +0000

The Token Ledger – 2026-08-06

Most cost‑impacting change: Qwen: Qwen3.6 27B saw a major price increase. Prompt rose from $0.289 to $0.60 per 1M tokens (+$0.311); completion rose from $2.40 to $3.60 per 1M tokens (+$1.20). Total cost per 1M tokens is now $4.20, up $1.511. Teams running large‑scale inference or fine‑tuning on this model should reassess budget allocations.

Added models

Meta: Muse Spark 1.2 – prompt $1.25 / 1M, completion $4.25 / 1M. Ideal for developers needing long‑context (1M tokens) generative tasks.
inclusionAI: Ling-3.0-flash – prompt $0.075 / 1M, completion $0.22 / 1M. Suitable for low‑latency, cost‑sensitive applications with moderate context (128K).

Price changes

MoonshotAI: Kimi K2.7 Code – prompt down $0.03 to $0.70 / 1M; completion unchanged at $3.50 / 1M. Minor savings for code‑generation workloads.
DeepSeek: DeepSeek V4 Flash 0423 – prompt down $0.0518 to $0.0882 / 1M; completion down $0.1036 to $0.1764 / 1M. Total reduction $0.1554 / 1M; beneficial for high‑volume flash inference.
Z.ai: GLM 5.1 – prompt down $0.014 to $0.952 / 1M; completion down $0.044 to $2.992 / 1M. Small overall cut ($0.058 / 1M) for balanced prompt/completion use.
MiniMax: MiniMax M2.5 – prompt up $0.07 to $0.22 / 1M; completion steady at $0.90 / 1M. Slight cost rise for users of this model.
Qwen: Qwen3 235B A22B Instruct 2507 – prompt down $0.0595 to $0.09 / 1M; completion down $0.048 to $0.55 / 1M. Total saving $0.1075 / 1M, relevant for large‑model deployments.

Cheapest models today (for reference): inclusionAI: Ling-2.6-flash ($0.01 / 1M prompt, $0.03 / 1M completion), IBM: Granite 4.0 Micro ($0.017 / 1M prompt, $0.112 / 1M completion), Mistral: Mistral Nemo ($0.019 / 1M prompt, $0.03 / 1M completion).

Originally published at The Token Ledger. Subscribe for the daily digest.

The Token Ledger – 2026-08-05

4663437Mehdi — Wed, 05 Aug 2026 09:47:02 +0000

The Token Ledger – 2026-08-05

MiniMax: MiniMax M2.7

What changed: Prompt and completion prices increased.
Numbers (per 1M tokens):
- Prompt: $0.25 → $0.27 (+$0.02)
- Completion: $1.00 → $1.08 (+$0.08)
Who should care: Teams running high‑volume completion workloads on MiniMax; the $0.08/M rise adds $80 per 1B tokens.

Mistral: Mistral Small 3.2 24B

What changed: Prompt and completion prices increased.
Numbers (per 1M tokens):
- Prompt: $0.075 → $0.09375 (+$0.01875)
- Completion: $0.20 → $0.25 (+$0.05)
Who should care: Users of Mistral’s 24B instruct model; the $0.05/M completion uplift adds $50 per 1B tokens.

No models were added or removed today. Total models tracked: 338.

Originally published at The Token Ledger. Subscribe for the daily digest.

The Token Ledger Digest – 2026-08-04

4663437Mehdi — Tue, 04 Aug 2026 09:49:18 +0000

The Token Ledger Digest – 2026-08-04

Z.ai: GLM 5.2 – Completion price fell from $3.74 to $2.42 per 1M tokens (‑$1.32/M). Prompt price dropped from $1.19 to $0.76 per 1M tokens (‑$0.43/M). Who should care: Teams running cost‑sensitive, long‑completion workloads on Z.ai.

Qwen: Qwen3.5‑122B‑A10B – Prompt ↓ $0.40→$0.26/M (‑$0.14/M). Completion ↓ $3.20→$2.08/M (‑$1.12/M). Who should care: Users of large‑scale Qwen models seeking lower generation cost.

Qwen: Qwen2.5 VL 72B Instruct – Prompt ↓ $0.80→$0.25/M (‑$0.55/M). Completion ↓ $1.00→$0.75/M (‑$0.25/M). Who should care: Vision‑language apps needing cheaper prompt processing.

MoonshotAI: Kimi K2.6 – Prompt ↓ $0.60→$0.589/M (‑$0.011/M). Completion ↓ $3.41→$2.48/M (‑$0.93/M). Who should care: Developers balancing prompt and completion costs.

Qwen: Qwen3.6 27B – Prompt ↓ $0.30→$0.289/M (‑$0.011/M). Completion ↑ $2.00→$2.40/M (+$0.40/M). Who should care: Slightly cheaper prompts but higher completion expense.

Meta: Llama 3.3 70B Instruct – Prompt ↓ $0.13→$0.10/M (‑$0.03/M). Completion ↓ $0.40→$0.32/M (‑$0.08/M). Who should care: General‑purpose Llama users seeing modest savings.

Qwen: Qwen3 Next 80B A3B Instruct – Prompt ↓ $0.10→$0.09/M (‑$0.01/M). Completion unchanged at $1.10/M. Who should care: Minor prompt‑cost reduction.

Qwen: Qwen3 Coder 30B A3B Instruct – Prompt unchanged $0.07/M. Completion ↓ $0.28→$0.27/M (‑$0.01/M). Who should care: Tiny completion‑cost trim.

Qwen: Qwen3 VL 30B A3B Instruct – Prompt ↑ $0.13→$0.15/M (+$0.02/M). Completion ↑ $0.52→$0.60/M (+$0.0. Who should care: Slight costlier.

*M0.08/M). *Who should care: Small increase for vision‑language tasks.

MythoMax 13B – Prompt ↑ $0.06→$0.08/M (+$0.02/M). Completion ↑ $0.06→$0.11/M (+$0.05/M). Who should care: Minor cost rise for this model.

Added: Qwen: Qwen3.8 Max – Prompt $2.00/M, Completion $6.00/M, 1M‑token context. Who should care: New high‑capacity option for enterprises needing massive context windows.

Originally published at The Token Ledger. Subscribe for the daily digest.

The Token Ledger Digest – 2026-08-03

4663437Mehdi — Mon, 03 Aug 2026 10:50:55 +0000

The Token Ledger Digest – 2026-08-03

Most cost‑impacting change: Z.ai’s GLM 5.2 saw a sharp price rise, raising both prompt and completion costs by roughly $0.90‑$2.85 per 1M tokens.

Price changes

Model	What changed	Old price (/1M)	New price (/1M)	Who should care
Z.ai: GLM 5.2	Prompt ↑ from $0.28 to $1.19; Completion ↑ from $0.89 to $3.74	$0.28 / $0.89	$1.19 / $3.74	Teams using GLM 5.2 for high‑volume generation; budget reassessment needed.
Qwen: Qwen3.5‑122B‑A10B	Prompt ↑ $0.26 → $0.40; Completion ↑ $2.08 → $3.20	$0.26 / $2.08	$0.40 / $3.20	Developers balancing quality vs cost for mid‑size LLMs.
Qwen: Qwen3 VL 235B A22B Thinking	Prompt ↑ $0.40 → $0.98; Completion ↓ $4.00 → $3.95	$0.40 / $4.00	$0.98 / $3.95	Vision‑language users; prompt cost up slightly, completion marginally cheaper.
Qwen: Qwen3 235B A22B Instruct 2507	Prompt ↑ $0.09 → $0.15; Completion ↑ $0.55 → $0.60	$0.09 / $0.55	$0.15 / $0.60	Cost‑sensitive applications using this instruct variant.
OpenAI: GPT‑5.6 Luna Pro	No meaningful change (prompt $0.10, completion $0.60 both unchanged)	$0.10 / $0.60	$0.10 / $0.60	No action required.
OpenAI: GPT‑5.6 Luna	Same as above	$0.10 / $0.60	$0.10 / $0.60	No action required.
OpenAI: GPT‑5.6 Terra Pro	No meaningful change (prompt $1.00, completion $6.00)	$1.00 / $6.00	$1.00 / $6.00	No action required.
OpenAI: GPT‑5.6 Terra	Same as above	$1.00 / $6.00	$1.00 / $6.00	No action required.

Per‑token prices were multiplied by 1,000,000 and rounded to two decimal places for readability.

Cheapest models today (per‑million tokens)

inclusionAI: Ling‑2.6‑flash – Prompt $0.01, Completion $0.03
Mistral: Mistral Nemo – Prompt $0.02, Completion $0.03
IBM: Granite 4.0 Micro – Prompt $0.02, Completion $0.11

No models were added or removed today. Total models tracked: 337.

Originally published at The Token Ledger. Subscribe for the daily digest.

The Token Ledger Digest – 2026-08-02

4663437Mehdi — Sun, 02 Aug 2026 09:19:16 +0000

The Token Ledger Digest – 2026-08-02

Price Drop – Z.ai: GLM 5.2

Prompt: fell from $0.76 /1M to $0.28 /1M (‑63%).
Completion: fell from $2.39 /1M to $0.89 /1M (‑63%).
Who should care: Teams running large‑scale inference or fine‑tuning on GLM 5.2 see per‑token costs cut by roughly two‑thirds.

Price Drop – DeepSeek: DeepSeek V4 Flash 0731

Prompt: dropped from $0.14 /1M to $0.09 /1M (‑36%).
Completion: dropped from $0.28 /1M to $0.18 /1M (‑36%).
Who should care: Users of the 0731 checkpoint benefit from lower latency‑cost trade‑offs for flash‑style workloads.

Price Increase – NVIDIA: Nemotron 3 Ultra

Prompt: rose from $0.50 /1M to $0.60 /1M (+20%).
Completion: rose from $2.20 /1M to $3.60 /1M (+64%).
Who should care: Cost‑sensitive applications relying on Nemotron 3 Ultra for completion‑heavy tasks will see noticeably higher bills.

Added Model – DeepSeek V4 Flash Latest

Prompt: $0.09 /1M
Completion: $0.18 /1M
Context: 1,048,576 tokens
Who should care: Developers needing ultra‑long context with flash‑level pricing can now access this newest variant.

Total models tracked: 337.

Originally published at The Token Ledger. Subscribe for the daily digest.

Token Ledger Digest – 2026-08-01

4663437Mehdi — Sat, 01 Aug 2026 09:19:05 +0000

Token Ledger Digest – 2026-08-01

Most impactful change

Z.ai: GLM 5.2 – Prompt price fell from $1.232 / 1M to $0.760 / 1M; completion price fell from $3.872 / 1M to $2.389 / 1M. Who should care: Cost‑sensitive applications using this model see ~48% lower prompt and ~38% lower completion costs.

Added

Thinking Machines: Inkling Small – New model with 524 k context, prompt $0.50 / 1M, completion $1.20 / 1M. Who should care: Teams needing long‑context generation at low cost.

Removed

30 batch models were deleted, including Google Gemini 3.6 Flash batch, Anthropic Claude Sonnet 5 batch, and OpenAI GPT‑5.5 batch. Their prompt prices ranged from $0.15 / 1M to $2.50 / 1M and completion prices from $1.25 / 1M to $25.00 / 1M. Who should care: Users relying on batch pricing must migrate to alternative endpoints or non‑batch versions.

Other price changes

NVIDIA: Nemotron 3 Ultra – Prompt $0.60 → $0.50 / 1M; completion $3.60 → $2.20 / 1M.
MoonshotAI: Kimi K2.6 – Prompt $0.95 → $0.60 / 1M; completion $4.00 → $3.41 / 1M.
Qwen: Qwen3 Coder 30B A3B Instruct – Prompt unchanged $0.07 / 1M; completion rose $0.27 → $0.28 / 1M.
Mistral: Mistral Small 3.2 24B – Prompt $0.10 → $0.075 / 1M; completion $0.30 → $0.20 / 1M.

Cheapest models today (for reference)

inclusionAI: Ling‑2.6‑flash – $0.01 / 1M prompt, $0.03 / 1M completion
IBM: Granite 4.0 Micro – $0.017 / 1M prompt, $0.112 / 1M completion
Mistral: Mistral Nemo – $0.019 / 1M prompt, $0.03 / 1M completion

Total models tracked: 336.

Originally published at The Token Ledger. Subscribe for the daily digest.

Token Ledger Digest – 2026-07-31

4663437Mehdi — Fri, 31 Jul 2026 09:52:25 +0000

Token Ledger Digest – 2026-07-31

Added

DeepSeek: DeepSeek V4 Flash 0731 – New model with 1M‑token context. Prompt $0.14/1M, completion $0.28/1M. Who should care: Teams needing long‑context generation at low cost.

Removed

OpenAI: o3 Deep Research – Prompt $10.00/1M, completion $40.00/1M.
OpenAI: o4 Mini Deep Research – Prompt $2.00/1M, completion $8.00/1M.
OpenAI: GPT-5 Codex – Prompt $1.25/1M, completion $10.00/1M. Who should care: Any workflows relying on these models must migrate to alternatives.

Price Changes

Model	Old Prompt ($/1M)	New Prompt ($/1M)	Δ Prompt	Old Compl. ($/1M)	New Compl. ($/1M)	Δ Compl.	Who should care
Poolside: Laguna S 2.1	0.10	0.09	–0.01	0.20	0.18	–0.02	Cost‑sensitive inference on Laguna S.
OpenAI: GPT-5.6 Luna Pro	0.50	0.10	–0.40	3.00	0.60	–2.40	Heavy users of Luna Pro see ~80% cost cut.
OpenAI: GPT-5.6 Luna	0.50	0.10	–0.40	3.00	0.60	–2.40	Same impact as Luna Pro.
OpenAI: GPT-5.6 Terra Pro	1.25	1.00	–0.25	7.50	6.00	–1.50	Terra Pro workloads become cheaper.
OpenAI: GPT-5.6 Terra	1.25	1.00	–0.25	7.50	6.00	–1.50	Same as Terra Pro.
Z.ai: GLM 5.2	0.70	1.23	+0.53	2.20	3.87	+1.67	GLM 5.2 now significantly more expensive.
NVIDIA: Nemotron 3 Ultra	0.50	0.60	+0.10	2.20	3.60	+1.40	Ultra‑scale users face higher spend.
MoonshotAI Kimi Latest	2.90	2.90	0.00	15.00	14.00	–1.00	Slight completion‑cost relief.
MoonshotAI: Kimi K2.6	0.65	0.95	+0.30	2.72	4.00	+1.28	Kimi K2.6 cost rises noticeably.
Qwen: Qwen3 VL 30B A3B Instruct	0.15	0.13	–0.02	0.60	0.52	–0.08	Minor savings for vision‑language tasks.
Qwen: Qwen3 235B A22B Thinking 2507	0.30	0.23	–0.07	3.00	2.30	–0.70	Thinking model becomes cheaper.

Most cost‑impacting change: The OpenAI GPT-5.6 Luna Pro and Luna models dropped prompt pricing from $0.50 to $0.10/1M and completion from $3.00 to $0.60/1M – an ~80% reduction, saving up to $2.80 per million tokens.

Cheapest Models Today

inclusionAI: Ling-2.6-flash – $0.01/1M prompt, $0.03/1M completion
IBM: Granite 4.0 Micro – $0.017/1M prompt, $0.112/1M completion
Mistral: Mistral Nemo – $0.019/1M prompt, $0.03/1M completion

Total models tracked: 365.

Originally published at The Token Ledger. Subscribe for the daily digest.

The Token Ledger Digest – 2026-07-30

4663437Mehdi — Thu, 30 Jul 2026 09:39:59 +0000

The Token Ledger Digest – 2026-07-30

Biggest cost impact: DeepSeek DeepSeek V3 – prompt price rose from $0.20 to $0.26 / 1M tokens (+$0.06), completion from $0.80 to $1.03 / 1M tokens (+$0.23). Total increase ≈ $0.29 / 1M tokens.

Who should care: Teams running high‑volume chat or code‑gen workloads on DeepSeek V3; budget forecasts need upward adjustment.

Other price changes:

Z.ai GLM 5.2 – prompt ↓ $0.74 → $0.70 / 1M (-$0.05); completion ↓ $2.34 → $2.20 / 1M (-$0.14). Total –$0.19 / 1M.

Who should care: Users of GLM 5.2 seeing modest savings.
MoonshotAI Kimi Latest – prompt ↓ $3.00 → $2.90 / 1M (-$0.10); completion unchanged at $15.00 / 1M. Total –$0.10 / 1M.

Who should care: Cost‑sensitive prompt‑heavy applications.
Google Gemma 4 31B – prompt ↓ $0.14 → $0.10 / 1M (-$0.04); completion ↓ $0.40 → $0.34 / 1M (-$0.06). Total –$0.10 / 1M.

Who should care: Developers using Gemma 4 for latency‑critical tasks.
Qwen Qwen3 Coder Next – prompt ↓ $0.18 → $0.12 / 1M (-$0.06); completion ↓ $0.90 → $0.80 / 1M (-$0.10). Total –$0.16 / 1M.

Who should care: Coding assistants benefiting from lower token cost.
Qwen Qwen3 VL 30B A3B Instruct – prompt ↑ $0.13 → $0.15 / 1M (+$0.02); completion ↑ $0.52 → $0.60 / 1M (+$0.08). Total +$0.10 / 1M.

Who should care: Vision‑language workloads seeing a slight cost rise.
Qwen Qwen2.5 7B Instruct – prompt ↑ $0.04 → $0.10 / 1M (+$0.06); completion ↑ $0.10 → $0.20 / 1M (+$0.10). Total +$0.16 / 1M.

Who should care: Light‑weight inference services now noticeably more expensive.

No models were added or removed today.

Cheapest models available: inclusionAI Ling‑2.6‑flash ($0.01 / 1M prompt, $0.03 / 1M completion), IBM Granite 4.0 Micro ($0.017 / 1M prompt, $0.112 / 1M completion), Mistral Nemo ($0.019 / 1M prompt, $0.03 / 1M completion).

Total models tracked: 367.

Originally published at The Token Ledger. Subscribe for the daily digest.

Token Ledger Digest – 2026-07-29

4663437Mehdi — Wed, 29 Jul 2026 09:48:30 +0000

Token Ledger Digest – 2026-07-29

The biggest cost impact today is a price cut for NVIDIA Nemotron 3 Ultra, dropping completion from $3.60 to $2.20 per 1M tokens (‑$1.40/1M).

Added (28 models)

Google Gemini 3.6 Flash (batch): added; prompt $0.75/1M, completion $3.75/1M; teams needing batch inference.
Google Gemini 3.5 Flash Lite (batch): added; prompt $0.15/1M, completion $1.25/1M; teams needing batch inference.
Anthropic Claude Sonnet 5 (batch): added; prompt $1.00/1M, completion $5.00/1M; teams needing batch inference.
Anthropic Claude Fable 5 (batch): added; prompt $5.00/1M, completion $25.00/1M; teams needing batch inference.
MiniMax MiniMax M3 (batch): added; prompt $0.15/1M, completion $0.60/1M; teams needing batch inference.
Anthropic Claude Opus 4.8 (batch): added; prompt $2.50/1M, completion $12.50/1M; teams needing batch inference.
Google Gemini 3.5 Flash (batch): added; prompt $0.75/1M, completion $4.50/1M; teams needing batch inference.
Google Gemini 3.1 Flash Lite (batch): added; prompt $0.125/1M, completion $0.75/1M; teams needing batch inference.
OpenAI GPT-5.5 (batch): added; prompt $2.50/1M, completion $15.00/1M; teams needing batch inference.
Anthropic Claude Opus 4.7 (batch): added; prompt $2.50/1M, completion $12.50/1M; teams needing batch inference.
OpenAI GPT-5.4 Nano (batch): added; prompt $0.10/1M, completion $0.625/1M; teams needing batch inference.
OpenAI GPT-5.4 Mini (batch): added; prompt $0.375/1M, completion $2.25/1M; teams needing batch inference.
OpenAI GPT-5.4 (batch): added; prompt $1.25/1M, completion $7.50/1M; teams needing batch inference.
Google Gemini 3.1 Pro Preview (batch): added; prompt $1.00/1M, completion $6.00/1M; teams needing batch inference.
Anthropic Claude Opus 4.6 (batch): added; prompt $2.50/1M, completion $12.50/1M; teams needing batch inference.

Removed (2 models)

Poolside Laguna M.1: removed; prompt $0.20/1M, completion $0.40/1M; users of this model.
Poolside Laguna M.1 (free): removed; prompt $0.00/1M, completion $0.00/1M; users of the free tier.

Price Changes (13 models)

Z.ai GLM 5.2: prompt ↓$0.7686→$0.7448/1M, completion ↓$2.4156→$2.3408/1M; developers using this model.
NVIDIA Nemotron 3 Ultra: prompt ↓$0.60→$0.50/1M, completion ↓$3.60→$2.20/1M; developers using this model.
Qwen Qwen3.6 Max Preview: prompt ↓$1.04→$1.027/1M, completion ↓$6.24→$6.162/1M; developers using this model.
Google Gemma 4 26B A4B: prompt ↓$0.14→$0.07/1M, completion ↓$0.42→$0.34/1M; developers using this model.
Qwen Qwen3 Coder Next: prompt ↑$0.11→$0.18/1M, completion ↑$0.80→$0.90/1M; developers using this model.
Qwen Qwen3 VL 8B Thinking: prompt ↑$0.117→$0.18/1M, completion ↑$1.365→$2.10/1M; developers using this model.
Qwen Qwen3 VL 30B A3B Thinking: prompt ↑$0.13→$0.20/1M, completion ↑$1.56→$2.40/1M; developers using this model.
Qwen Qwen3 VL 30B A3B Instruct: prompt ↓$0.15→$0.13/1M, completion ↓$0.60→$0.52/1M; developers using this model.
Qwen Qwen3 VL 235B A22B Thinking: prompt ↑$0.26→$0.40/1M, completion ↑$2.60→$4.00/1M; developers using this model.
Qwen Qwen3 Next 80B A3B Thinking: prompt ↑$0.0975→$0.15/1M, completion ↑$0.78→$1.20/1M; developers using this model.
Qwen Qwen Plus 0728 (thinking): prompt ↑$0.26→$0.40/1M, completion ↑$0.78→$1.20/1M; developers using this model.
Qwen Qwen3 30B A3B Thinking 2507: prompt ↑$0.13→$0.20/1M, completion ↑$1.56→$2.40/1M; developers using this model.
OpenAI gpt-oss-20b: prompt unchanged $0.03/1M, completion ↓$0.14→$0.13/1M; developers using this model.

Cheapest models today: inclusionAI Ling-2.6-flash ($0.01/1M prompt, $0.03/1M completion), IBM Granite 4.0 Micro ($0.017/1M prompt, $0.112/1M completion), Mistral Nemo ($0.019/1M prompt, $0.03/1M completion).

Originally published at The Token Ledger. Subscribe for the daily digest.

AI Model Pricing Digest – 2026-07-28

4663437Mehdi — Tue, 28 Jul 2026 09:46:27 +0000

AI Model Pricing Digest – 2026-07-28

Most cost‑impacting change: OpenAI cut prices for the GPT‑5.6 Terra series, slashing both prompt and completion costs by ~40‑50%.

Added

Qwen: Qwen3.7 Flash – New model. Prompt $0.03/1M, completion $0.13/1M. Who should care: Teams needing ultra‑long context (1M tokens) at low cost.

Removed

OpenAI: GPT-5 Chat – Deleted. Was $1.25/1M prompt, $10/1M completion. Who should care: Users of this model must migrate to alternatives.
OpenAI: GPT-4o Search Preview – Deleted. Was $2.5/1M prompt, $10/1M completion. Who should care: Anyone relying on search‑augmented preview should switch.

Price Changes

OpenAI: GPT-5.6 Luna Pro – Prompt ↓ $1.00 → $0.50/1M (-$0.50); Completion ↓ $6.00 → $3.00/1M (-$3.00). Who should care: Cost‑sensitive workloads using Luna Pro.
OpenAI: GPT-5.6 Luna – Same cuts as Luna Pro. Who should care: Luna users.
OpenAI: GPT-5.6 Terra Pro – Prompt ↓ $2.50 → $1.25/1M (-$1.25); Completion ↓ $15.00 → $7.50/1M (-$7.50). Who should care: High‑volume Terra Pro users see ~40% savings.
OpenAI: GPT-5.6 Terra – Identical cuts to Terra Pro. Who should care: Terra users.
Z.ai: GLM 5.2 – Prompt ↓ $0.8036 → $0.7686/1M (-$0.035); Completion ↓ $2.5256 → $2.4156/1M (-$0.11). Who should care: Minor savings for GLM 5.2 adopters.
NVIDIA: Nemotron 3 Ultra – Prompt ↑ $0.50 → $0.60/1M (+$0.10); Completion ↑ $2.20 → $3.60/1M (+$1.40). Who should care: Budget impact for Nemotron 3 Ultra users.
Google: Gemma 4 26B A4B – Prompt ↑ $0.12 → $0.14/1M (+$0.02); Completion ↑ $0.35 → $0.42/1M (+$0.07). Who should care: Slight cost rise for Gemma 4 users.
Qwen: Qwen3.5-35B-A3B – Prompt ↓ $0.15 → $0.14/1M (-$0.01); Completion unchanged $1.00/1M. Who should care: Negligible saving for this model.

Total models tracked: 341.

Originally published at The Token Ledger. Subscribe for the daily digest.

Token Ledger Digest – 2026-07-27

4663437Mehdi — Mon, 27 Jul 2026 10:50:17 +0000

Token Ledger Digest – 2026-07-27

Most cost‑impacting change

NVIDIA: Nemotron 3 Ultra – completion price dropped from $3.60 to $2.20 per 1M tokens; prompt price fell from $0.60 to $0.50 per 1M tokens. Who should care: Existing users see a ~39% reduction in generation cost; new adopters benefit from lower pricing.

Other price changes

Z.ai: GLM 5.2 – prompt rose from $0.68 to $0.80 per 1M; completion rose from $2.15 to $2.53 per 1M. Who should care: Sizable cost increase for workloads using GLM 5.2.
MoonshotAI: Kimi K2.7 Code – prompt fell from $0.75 to $0.73 per 1M; completion unchanged at $3.50 per 1M. Who should care: Minor prompt‑side savings.
Qwen: Qwen3.5-35B-A3B – prompt rose from $0.14 to $0.15 per 1M; completion unchanged at $1.00 per 1M. Who should care: Negligible impact.

Removed models

OpenAI: GPT-4o-mini Search Preview – prompt $0.15/1M, completion $0.60/1M.
Inflection: Inflection 3 Pi – prompt $2.50/1M, completion $10.00/1M.
Inflection: Inflection 3 Productivity – same pricing as Pi. Who should care: Anyone relying on these models must migrate to alternatives; they are no longer available for new calls.

Total models tracked: 342.

Originally published at The Token Ledger. Subscribe for the daily digest.

The Token Ledger – 2026-07-26

4663437Mehdi — Sun, 26 Jul 2026 09:22:29 +0000

The Token Ledger – 2026-07-26

Most cost‑impacting change: Qwen’s Qwen3.6 27B completion price fell from $2.40 to $2.00 per 1M tokens (‑$0.40/M), a 16.7% reduction.

Price changes

Z.ai: GLM 5.2
- Prompt: $0.749 → $0.685 /M
- Completion: $2.354 → $2.152 /M
- Who should care: Teams running latency‑critical inference where both prompt and completion costs matter.
MoonshotAI: Kimi K2.7 Code
- Prompt: $0.78 → $0.75 /M
- Completion: unchanged at $3.50 /M
- Who should care: Code‑generation workflows sensitive to prompt pricing.
Qwen: Qwen3.6 27B (lead change)
- Prompt: $0.289 → $0.300 /M (+$0.011/M)
- Completion: $2.400 → $2.000 /M (‑$0.400/M)
- Who should care: Applications generating long completions (e.g., chat, summarization) that benefit from lower output cost.
DeepSeek: DeepSeek V4 Flash
- Prompt: $0.094 → $0.14 /M (+$0.046/M)
- Completion: $0.188 → $0.28 /M (+$0.092/M)
- Who should care: Cost‑conscious users of flash‑mode models; higher per‑token cost may shift usage to alternatives.
OpenAI: gpt‑oss‑20b
- Prompt: unchanged at $0.030 /M
- Completion: $0.130 → $0.140 /M (+$0.010/M)
- Who should care: Light‑weight completion tasks where a small cost increase may affect budgeting.
Qwen: Qwen3 30B A3B Instruct 2507
- Prompt: $0.100 → $0.048 /M (‑$0.052/M)
- Completion: $0.300 → $0.193 /M (‑$0.107/M)
- Who should care: Instruction‑following pipelines seeing notable savings on both input and output.
Qwen: Qwen3 30B A3B
- Prompt: $0.130 → $0.120 /M (‑$0.010/M)
- Completion: $0.520 → $0.500 /M (‑$0.020/M)
- Who should care: General‑purpose deployments where modest per‑token reductions accumulate over high volume.

No models were added or removed today.

Three cheapest models (per‑token pricing):

inclusionAI: Ling‑2.6‑flash – Prompt $0.00001, Completion $0.00003 /M
IBM: Granite 4.0 Micro – Prompt $0.000017, Completion $0.000112 /M
Mistral: Mistral Nemo – Prompt $0.000019, Completion $0.00003 /M

Originally published at The Token Ledger. Subscribe for the daily digest.