You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Detect when the "Google Cloud Credentials" Vertex field has been given a filesystem path instead of the raw JSON contents of a service-account key file. The runtime now skips JSON.parse for path-shaped input, logs a single specific console warning naming the sibling "Google Cloud Key File Path" field and `GOOGLE_APPLICATION_CREDENTIALS` env var, and the settings UI shows an inline warning under the field while the input still looks like a path. Auth behavior is unchanged for correctly-configured users.
"Llama 4 Scout is a smaller, faster variant of Llama 4 with multimodal capabilities, ideal for quick iterations and cost-effective deployments.",
242
245
},
246
+
"accounts/fireworks/models/kimi-k2p6": {
247
+
maxTokens: 16384,
248
+
contextWindow: 262144,
249
+
supportsImages: true,
250
+
supportsPromptCache: true,
251
+
inputPrice: 0.95,
252
+
outputPrice: 4.0,
253
+
cacheReadsPrice: 0.16,
254
+
description:
255
+
"Kimi K2.6 is Moonshot AI's latest flagship agentic model, building on K2.5 with stronger long-horizon reasoning, multi-step tool use, and unified vision/text understanding.",
256
+
},
257
+
"accounts/fireworks/models/deepseek-v4-pro": {
258
+
maxTokens: 16384,
259
+
contextWindow: 1048576,
260
+
supportsImages: false,
261
+
supportsPromptCache: true,
262
+
inputPrice: 1.74,
263
+
outputPrice: 3.48,
264
+
cacheReadsPrice: 0.14,
265
+
description:
266
+
"DeepSeek V4 Pro is the latest iteration of the DeepSeek model family, with improved reasoning, code generation, and instruction following over the V3 series.",
267
+
},
268
+
"accounts/fireworks/models/glm-5p1": {
269
+
maxTokens: 25344,
270
+
contextWindow: 202752,
271
+
supportsImages: false,
272
+
supportsPromptCache: true,
273
+
inputPrice: 1.4,
274
+
outputPrice: 4.4,
275
+
cacheReadsPrice: 0.26,
276
+
description:
277
+
"Z.ai GLM-5.1 is the latest coding-focused model in the GLM family, with exceptional performance on complex programming tasks and enhanced reasoning and code generation quality.",
0 commit comments