Anuncio

Agendator

Reservas online gratis para tu barbería o salón.

Página de reservas, agenda y recordatorios por email. Sin cookies ni anuncios para tus clientes.

Crear mi página gratis

DeepSeek V4.1 Flash 1017 output price falls to $0.15 per million tokens

660 palabras

Por ahora este informe solo está disponible en inglés.

DeepSeek V4.1 Flash 1017 cut its output price from $0.50 to $0.15 per million tokens, the largest single output reduction among the 12 price changes recorded on OpenRouter for 2026-11-19. The same cut applies to DeepSeek Flash Latest. Across 480 tracked models, 17 are free today, 189 are open weights, and two new Google Gemini 4 Neon listings arrived with no models removed.

DeepSeek cuts output, raises input

DeepSeek V4.1 Flash 1017 moved input from $0.036 to $0.035 per million tokens and output from $0.50 to $0.15. DeepSeek Flash Latest shows the identical pair of moves. DeepSeek V4 Flash Latest went the other way on input, from $0.0434 to $0.047, with output unchanged at $0.30.

DeepSeek R1 0528 raised both sides: input from $0.50 to $0.571 and output from $2.15 to $2.286 per million tokens.

Qwen moves in both directions

Qwen3.5-9B edged down, with input from $0.149 to $0.142 and output from $0.22 to $0.21. Qwen3.5-27B raised input from $0.26 to $0.30 and cut output from $2.6 to $2.4. Qwen3 Next 80B A3B Instruct raised input from $0.09 to $0.10 with output held at $1.1.

Qwen3.8 27B cut input from $0.15 to $0.06 and raised output from $1.875 to $2.2 per million tokens.

Other price changes

Google Gemma 3 27B held input at $0.08 and raised output from $0.35 to $0.45. MoonshotAI Kimi K2.7 Code cut input from $0.7 to $0.6 and output from $3.75 to $3.5.

Z.ai GLM 5.3 Flash raised input from $0.0399 to $0.15 and output from $0.0784 to $0.5. Xiaomi MiMo-V2.5-Pro raised input from $0.435 to $0.48 and output from $0.87 to $1.8 per million tokens.

New listings and the free tier

Google added Gemini 4 Neon at $3.00 input and $15.00 output per million tokens, and Gemini 4 Neon (batch) at $1.50 input and $7.50 output, both with 1048576 context tokens and closed weights. No models were removed, and no model became free or stopped being free.

The 17 free models include inclusionAI Ling 3.2 Flash (free), Particle AI Particle-1 (free), Thinking Machines Inkling 2 Small (free), Baidu ERNIE-5.5-Flash-Preview (free), four NVIDIA Nemotron listings, Google Lyria 3 Pro Preview and Lyria 3 Clip Preview, Google Gemma 4 26B A4B (free) and Gemma 4 31B (free), Dots Studio Dots3-Note Preview (free), LiquidAI LFM3-1.7B (free), Recursive Flint (free), Cohere North Mini Code (free), and Free Models Router. Most carry 1048576 context tokens; Baidu lists 1024000, the NVIDIA entries 1000000, Dots Studio 512000, and LiquidAI, Recursive and both Gemma 4 entries 262144.

Cheapest paid options

By input price, inclusionAI Ling 3.0 Tiny leads at $0.015 per million tokens with output at $0.045 and 262144 context tokens. OpenAI gpt-oss-20b follows at $0.018 input and $0.09 output, then Meta Llama 3.1 8B Instruct at $0.02 input and $0.04 output.

inclusionAI Ling 3.0 Flash is listed at $0.021 input and $0.063 output, Z.ai GLM 5.3 Flash (batch) at $0.024 input and $0.069 output with 1048576 context tokens, OpenAI gpt-oss-20b (batch) at $0.024 input and $0.112 output, Nex AGI Nex-N2.5-Mini at $0.025 input and $0.1 output, and OpenAI GPT-5 Nano (batch) at $0.025 input and $0.2 output with 400000 context tokens.

Context windows

The largest context windows on the board are 2000000 tokens, held by SpaceXAI Grok 4.20 Multi-Agent, SpaceXAI Grok 4.20, Auto Router, Auto Code and Auto Router (Beta).

  • DeepSeek V4.1 Flash 1017 and DeepSeek Flash Latest both cut output from $0.50 to $0.15 per million tokens while input moved from $0.036 to $0.035.
  • DeepSeek R1 0528 raised input to $0.571 and output to $2.286 per million tokens.
  • Z.ai GLM 5.3 Flash raised input from $0.0399 to $0.15 and output from $0.0784 to $0.5 per million tokens.
  • Google Gemini 4 Neon entered at $3.00 input and $15.00 output per million tokens, with a batch listing at $1.50 and $7.50.
  • inclusionAI Ling 3.0 Tiny is the cheapest paid model by input at $0.015 per million tokens, with output at $0.045.

Ver los datos →

DeepSeek V4.1 Flash 1017 output price falls to $0.15 per million tokens