← All models
讯飞星火 X2.6

iFlytek Spark X2.6

1 providers

Spark X2.6 is iFlytek's flagship deep-reasoning LLM released in October 2026 — a 293B-parameter MoE model (~30B activated per token) trained entirely on domestic compute. It offers a 1M-token context window, visible chain-of-thought reasoning, and significantly improved agentic and coding capabilities. Available on the Xingchen MaaS platform at a limited-time 40%-off price (¥0.9 input / ¥3.6 output per 1M tokens), with OpenAI Chat, OpenAI Responses and Anthropic-compatible APIs — plugs straight into Claude Code, Codex and other agent tools.

NewFreeLimited-time PriceReasoningAgentLong ContextCodeiFlytek
iFlytek's flagship deep-reasoning model: 293B MoE, 1M context, visible chain-of-thought, stronger coding and agent capabilities.

Compare providers

Same underlying model, served by different providers. Compare pricing, context and latency, then call via the exact model_id shown.

ProviderInput / MOutput / MContextLatencyStatusmodel_idDetail
iFlytek SparkiFlytek Spark X2.6Lowest
¥0.90¥3.61000K2287msAvailablespark-x2.6View →